Add files using upload-large-folder tool

flat/objects/00/03/00036db0-78b4-448b-b651-180858272805.json

@@ -0,0 +1,448 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "CoQuIR/Instructor-large/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Mohamed bin Zayed University of Artificial Intelligence",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Instructor-large",
+    "name": "Instructor-large",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the average rank-based margin between code using modern APIs (positive) and deprecated APIs (negative) on the DepreAPI dataset. This task is challenging for most models, with Voyage-code-3 showing a clear advantage in prioritizing maintainable code.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MRS (%) - Maintainability (DepreAPI)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Margin-based Ranking Score for Maintainability on DepreAPI"
+        },
+        "metric_id": "margin_based_ranking_score_for_maintainability_on_depreapi",
+        "metric_name": "Margin-based Ranking Score for Maintainability on DepreAPI",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.97
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#margin_based_ranking_score_for_maintainability_on_depreapi"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the average rank-based margin between correct (positive) and buggy (negative) code snippets on the Defects4J dataset. MRS is stricter than PPA, quantifying how much higher correct code is ranked. A score near 0 indicates no quality awareness.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MRS (%) - Correctness (Defects4J)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Margin-based Ranking Score for Correctness on Defects4J"
+        },
+        "metric_id": "margin_based_ranking_score_for_correctness_on_defects4j",
+        "metric_name": "Margin-based Ranking Score for Correctness on Defects4J",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 4.43
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#margin_based_ranking_score_for_correctness_on_defects4j"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the average rank-based margin between efficient (positive) and inefficient (negative) code snippets on the CodeNet-E dataset. MRS is stricter than PPA, quantifying how much higher efficient code is ranked. A score near 0 indicates no quality awareness.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MRS (%) - Efficiency (CodeNet-E)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Margin-based Ranking Score for Efficiency on CodeNet-E"
+        },
+        "metric_id": "margin_based_ranking_score_for_efficiency_on_codenet_e",
+        "metric_name": "Margin-based Ranking Score for Efficiency on CodeNet-E",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": -0.52
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#margin_based_ranking_score_for_efficiency_on_codenet_e"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the average rank-based margin between efficient (positive) and inefficient (negative) SQL queries on the SQLR2 dataset. MRS is stricter than PPA, quantifying how much higher efficient queries are ranked. A score near 0 indicates no quality awareness.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MRS (%) - Efficiency (SQLR2)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Margin-based Ranking Score for SQL Query Efficiency on SQLR2"
+        },
+        "metric_id": "margin_based_ranking_score_for_sql_query_efficiency_on_sqlr2",
+        "metric_name": "Margin-based Ranking Score for SQL Query Efficiency on SQLR2",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 12.64
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#margin_based_ranking_score_for_sql_query_efficiency_on_sqlr2"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the average rank-based margin between patched, secure (positive) and vulnerable (negative) code snippets on the CVEFixes dataset. A score near 0 indicates little to no awareness of security fixes.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MRS (%) - Security (CVEFixes)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Margin-based Ranking Score for Security on CVEFixes"
+        },
+        "metric_id": "margin_based_ranking_score_for_security_on_cvefixes",
+        "metric_name": "Margin-based Ranking Score for Security on CVEFixes",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.05
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#margin_based_ranking_score_for_security_on_cvefixes"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the average rank-based margin between secure (positive) and insecure (negative) code snippets on the SafeCoder dataset. Many models score below zero, indicating a preference for insecure code on this challenging task.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MRS (%) - Security (SafeCoder)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Margin-based Ranking Score for Security on SafeCoder"
+        },
+        "metric_id": "margin_based_ranking_score_for_security_on_safecoder",
+        "metric_name": "Margin-based Ranking Score for Security on SafeCoder",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.98
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#margin_based_ranking_score_for_security_on_safecoder"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the model's ability to prefer bug-free code over incorrect counterparts on the CodeNet-B dataset. PPA is the proportion of positive-negative pairs where the model assigns a higher score to the positive (correct) sample. A score of 50% is random chance.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PPA (%) - Correctness (CodeNet-B)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Pairwise Preference Accuracy for Correctness on CodeNet-B"
+        },
+        "metric_id": "pairwise_preference_accuracy_for_correctness_on_codenet_b",
+        "metric_name": "Pairwise Preference Accuracy for Correctness on CodeNet-B",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 43.3
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#pairwise_preference_accuracy_for_correctness_on_codenet_b"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the model's ability to prefer bug-free code over buggy counterparts sourced from real Java projects in the Defects4J dataset. PPA is the proportion of positive-negative pairs where the model assigns a higher score to the positive (correct) sample.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PPA (%) - Correctness (Defects4J)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Pairwise Preference Accuracy for Correctness on Defects4J"
+        },
+        "metric_id": "pairwise_preference_accuracy_for_correctness_on_defects4j",
+        "metric_name": "Pairwise Preference Accuracy for Correctness on Defects4J",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 57.64
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#pairwise_preference_accuracy_for_correctness_on_defects4j"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the model's ability to prefer efficient code over functionally equivalent but suboptimal implementations on the CodeNet-E dataset. PPA is the proportion of positive-negative pairs where the model assigns a higher score to the positive (efficient) sample.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PPA (%) - Efficiency (CodeNet-E)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Pairwise Preference Accuracy for Efficiency on CodeNet-E"
+        },
+        "metric_id": "pairwise_preference_accuracy_for_efficiency_on_codenet_e",
+        "metric_name": "Pairwise Preference Accuracy for Efficiency on CodeNet-E",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 42.67
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#pairwise_preference_accuracy_for_efficiency_on_codenet_e"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the model's ability to prefer efficient SQL queries over inefficient ones on the SQLR2 dataset. PPA is the proportion of positive-negative pairs where the model assigns a higher score to the positive (efficient) sample.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PPA (%) - Efficiency (SQLR2)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Pairwise Preference Accuracy for SQL Query Efficiency on SQLR2"
+        },
+        "metric_id": "pairwise_preference_accuracy_for_sql_query_efficiency_on_sqlr2",
+        "metric_name": "Pairwise Preference Accuracy for SQL Query Efficiency on SQLR2",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 70.12
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#pairwise_preference_accuracy_for_sql_query_efficiency_on_sqlr2"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the model's ability to prefer code using modern, recommended APIs over code that relies on deprecated constructs on the DepreAPI dataset. PPA is the proportion of positive-negative pairs where the model assigns a higher score to the positive (updated) sample.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PPA (%) - Maintainability (DepreAPI)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Pairwise Preference Accuracy for Maintainability on DepreAPI"
+        },
+        "metric_id": "pairwise_preference_accuracy_for_maintainability_on_depreapi",
+        "metric_name": "Pairwise Preference Accuracy for Maintainability on DepreAPI",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 49.98
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#pairwise_preference_accuracy_for_maintainability_on_depreapi"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the model's ability to prefer patched, secure code over code with known vulnerabilities from the CVEFixes dataset. PPA is the proportion of positive-negative pairs where the model assigns a higher score to the positive (secure) sample.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PPA (%) - Security (CVEFixes)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Pairwise Preference Accuracy for Security on CVEFixes"
+        },
+        "metric_id": "pairwise_preference_accuracy_for_security_on_cvefixes",
+        "metric_name": "Pairwise Preference Accuracy for Security on CVEFixes",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 54.39
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#pairwise_preference_accuracy_for_security_on_cvefixes"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the average rank-based margin between correct (positive) and incorrect (negative) code snippets on the CodeNet-B dataset. MRS is stricter than PPA, quantifying how much higher correct code is ranked. A score near 0 indicates no quality awareness.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MRS (%) - Correctness (CodeNet-B)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Margin-based Ranking Score for Correctness on CodeNet-B"
+        },
+        "metric_id": "margin_based_ranking_score_for_correctness_on_codenet_b",
+        "metric_name": "Margin-based Ranking Score for Correctness on CodeNet-B",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": -0.73
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#margin_based_ranking_score_for_correctness_on_codenet_b"
+    },
+    {
+      "evaluation_name": "CoQuIR",
+      "source_data": {
+        "dataset_name": "CoQuIR",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/CoQuIR"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the model's ability to prefer secure code over implementations with insecure patterns on the SafeCoder dataset. PPA is the proportion of positive-negative pairs where the model assigns a higher score to the positive (secure) sample.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PPA (%) - Security (SafeCoder)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Pairwise Preference Accuracy for Security on SafeCoder"
+        },
+        "metric_id": "pairwise_preference_accuracy_for_security_on_safecoder",
+        "metric_name": "Pairwise Preference Accuracy for Security on SafeCoder",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 51.07
+      },
+      "evaluation_result_id": "CoQuIR/Instructor-large/1771591481.616601#coquir#pairwise_preference_accuracy_for_security_on_safecoder"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/04/00042efa-06d0-4a41-a092-c43b666f2583.json

@@ -0,0 +1,88 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "CiteEval/LongCite-8B/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Google",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "LongCite-8B",
+    "name": "LongCite-8B",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "CiteEval",
+      "source_data": {
+        "dataset_name": "CiteEval",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.01829"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the citation quality of various large language models using the CITEEVAL-AUTO metric on the CiteBench test set. The 'Full' scenario assesses all statements that require a citation, penalizing models for missing citations. This provides a comprehensive measure of both accuracy and completeness of source attribution. The score is normalized, with higher values being better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Citation Quality Score (Full Scenario)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Citation Quality on CiteBench (Full Scenario)"
+        },
+        "metric_id": "citation_quality_on_citebench_full_scenario",
+        "metric_name": "Citation Quality on CiteBench (Full Scenario)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.559
+      },
+      "evaluation_result_id": "CiteEval/LongCite-8B/1771591481.616601#citeeval#citation_quality_on_citebench_full_scenario"
+    },
+    {
+      "evaluation_name": "CiteEval",
+      "source_data": {
+        "dataset_name": "CiteEval",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.01829"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the citation quality of various large language models using the CITEEVAL-AUTO metric on the CiteBench test set. The 'Cited' scenario assesses only the statements that already have a citation, ignoring uncited statements. This metric focuses on the accuracy and relevance of the provided citations, rather than their completeness. The score is normalized, with higher values being better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Citation Quality Score (Cited Scenario)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Citation Quality on CiteBench (Cited Scenario)"
+        },
+        "metric_id": "citation_quality_on_citebench_cited_scenario",
+        "metric_name": "Citation Quality on CiteBench (Cited Scenario)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.846
+      },
+      "evaluation_result_id": "CiteEval/LongCite-8B/1771591481.616601#citeeval#citation_quality_on_citebench_cited_scenario"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/08/000806c2-ab3a-4ae3-88ef-9f56124d51f4.json

@@ -0,0 +1,304 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "vals-ai/lcb/minimax_MiniMax-M2.1/1777395187.3170502",
+  "retrieved_timestamp": "1777395187.3170502",
+  "source_metadata": {
+    "source_name": "Vals.ai Leaderboard - LiveCodeBench",
+    "source_type": "documentation",
+    "source_organization_name": "Vals.ai",
+    "source_organization_url": "https://www.vals.ai",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "benchmark_slug": "lcb",
+      "benchmark_name": "LiveCodeBench",
+      "benchmark_updated": "2026-04-21",
+      "dataset_type": "public",
+      "industry": "coding",
+      "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb",
+      "extraction_method": "static_astro_benchmark_view_props"
+    }
+  },
+  "eval_library": {
+    "name": "Vals.ai",
+    "version": "unknown"
+  },
+  "model_info": {
+    "name": "MiniMax-M2.1",
+    "id": "minimax/MiniMax-M2.1",
+    "developer": "minimax",
+    "additional_details": {
+      "vals_model_id": "minimax/MiniMax-M2.1",
+      "vals_provider": "MiniMax"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_result_id": "lcb:easy:minimax/MiniMax-M2.1:score",
+      "evaluation_name": "vals_ai.lcb.easy",
+      "source_data": {
+        "dataset_name": "LiveCodeBench - Easy",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/lcb"
+        ],
+        "additional_details": {
+          "benchmark_slug": "lcb",
+          "task_key": "easy",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Easy).",
+        "metric_id": "vals_ai.lcb.easy.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb"
+        }
+      },
+      "score_details": {
+        "score": 96.273,
+        "details": {
+          "benchmark_slug": "lcb",
+          "benchmark_name": "LiveCodeBench",
+          "benchmark_updated": "2026-04-21",
+          "task_key": "easy",
+          "task_name": "Easy",
+          "dataset_type": "public",
+          "industry": "coding",
+          "raw_score": "96.273",
+          "raw_stderr": "1.056",
+          "latency": "36.87",
+          "cost_per_test": "0.002892",
+          "temperature": "1",
+          "top_p": "0.95",
+          "max_output_tokens": "131000",
+          "provider": "MiniMax"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.056,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 1.0,
+          "top_p": 0.95,
+          "max_tokens": 131000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "lcb:hard:minimax/MiniMax-M2.1:score",
+      "evaluation_name": "vals_ai.lcb.hard",
+      "source_data": {
+        "dataset_name": "LiveCodeBench - Hard",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/lcb"
+        ],
+        "additional_details": {
+          "benchmark_slug": "lcb",
+          "task_key": "hard",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Hard).",
+        "metric_id": "vals_ai.lcb.hard.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb"
+        }
+      },
+      "score_details": {
+        "score": 62.571,
+        "details": {
+          "benchmark_slug": "lcb",
+          "benchmark_name": "LiveCodeBench",
+          "benchmark_updated": "2026-04-21",
+          "task_key": "hard",
+          "task_name": "Hard",
+          "dataset_type": "public",
+          "industry": "coding",
+          "raw_score": "62.571",
+          "raw_stderr": "2.587",
+          "latency": "529.6",
+          "cost_per_test": "0.038707",
+          "temperature": "1",
+          "top_p": "0.95",
+          "max_output_tokens": "131000",
+          "provider": "MiniMax"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.587,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 1.0,
+          "top_p": 0.95,
+          "max_tokens": 131000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "lcb:medium:minimax/MiniMax-M2.1:score",
+      "evaluation_name": "vals_ai.lcb.medium",
+      "source_data": {
+        "dataset_name": "LiveCodeBench - Medium",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/lcb"
+        ],
+        "additional_details": {
+          "benchmark_slug": "lcb",
+          "task_key": "medium",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Medium).",
+        "metric_id": "vals_ai.lcb.medium.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb"
+        }
+      },
+      "score_details": {
+        "score": 86.423,
+        "details": {
+          "benchmark_slug": "lcb",
+          "benchmark_name": "LiveCodeBench",
+          "benchmark_updated": "2026-04-21",
+          "task_key": "medium",
+          "task_name": "Medium",
+          "dataset_type": "public",
+          "industry": "coding",
+          "raw_score": "86.423",
+          "raw_stderr": "1.75",
+          "latency": "171.746",
+          "cost_per_test": "0.013857",
+          "temperature": "1",
+          "top_p": "0.95",
+          "max_output_tokens": "131000",
+          "provider": "MiniMax"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.75,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 1.0,
+          "top_p": 0.95,
+          "max_tokens": 131000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "lcb:overall:minimax/MiniMax-M2.1:score",
+      "evaluation_name": "vals_ai.lcb.overall",
+      "source_data": {
+        "dataset_name": "LiveCodeBench - Overall",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/lcb"
+        ],
+        "additional_details": {
+          "benchmark_slug": "lcb",
+          "task_key": "overall",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for LiveCodeBench (Overall).",
+        "metric_id": "vals_ai.lcb.overall.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/lcb"
+        }
+      },
+      "score_details": {
+        "score": 81.756,
+        "details": {
+          "benchmark_slug": "lcb",
+          "benchmark_name": "LiveCodeBench",
+          "benchmark_updated": "2026-04-21",
+          "task_key": "overall",
+          "task_name": "Overall",
+          "dataset_type": "public",
+          "industry": "coding",
+          "raw_score": "81.756",
+          "raw_stderr": "1.099",
+          "latency": "246.072",
+          "cost_per_test": "0.018485",
+          "temperature": "1",
+          "top_p": "0.95",
+          "max_output_tokens": "131000",
+          "provider": "MiniMax"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.099,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 1.0,
+          "top_p": 0.95,
+          "max_tokens": 131000,
+          "max_attempts": 1
+        }
+      }
+    }
+  ]
+}

flat/objects/00/0c/000c2d01-2cf9-4c19-8118-3a29367de6c4.json

@@ -0,0 +1,96 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "vals-ai/case_law_v2/anthropic_claude-opus-4-5-20251101-thinking/1777395187.3170502",
+  "retrieved_timestamp": "1777395187.3170502",
+  "source_metadata": {
+    "source_name": "Vals.ai Leaderboard - CaseLaw (v2)",
+    "source_type": "documentation",
+    "source_organization_name": "Vals.ai",
+    "source_organization_url": "https://www.vals.ai",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "benchmark_slug": "case_law_v2",
+      "benchmark_name": "CaseLaw (v2)",
+      "benchmark_updated": "2026-04-21",
+      "dataset_type": "private",
+      "industry": "legal",
+      "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2",
+      "extraction_method": "static_astro_benchmark_view_props"
+    }
+  },
+  "eval_library": {
+    "name": "Vals.ai",
+    "version": "unknown"
+  },
+  "model_info": {
+    "name": "claude-opus-4-5-20251101-thinking",
+    "id": "anthropic/claude-opus-4-5-20251101-thinking",
+    "developer": "anthropic",
+    "additional_details": {
+      "vals_model_id": "anthropic/claude-opus-4-5-20251101-thinking",
+      "vals_provider": "Anthropic"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_result_id": "case_law_v2:overall:anthropic/claude-opus-4-5-20251101-thinking:score",
+      "evaluation_name": "vals_ai.case_law_v2.overall",
+      "source_data": {
+        "dataset_name": "CaseLaw (v2) - Overall",
+        "source_type": "other",
+        "additional_details": {
+          "benchmark_slug": "case_law_v2",
+          "task_key": "overall",
+          "dataset_type": "private",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for CaseLaw (v2) (Overall).",
+        "metric_id": "vals_ai.case_law_v2.overall.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/case_law_v2"
+        }
+      },
+      "score_details": {
+        "score": 62.594,
+        "details": {
+          "benchmark_slug": "case_law_v2",
+          "benchmark_name": "CaseLaw (v2)",
+          "benchmark_updated": "2026-04-21",
+          "task_key": "overall",
+          "task_name": "Overall",
+          "dataset_type": "private",
+          "industry": "legal",
+          "raw_score": "62.594",
+          "raw_stderr": "0.107",
+          "latency": "16.819",
+          "cost_per_test": "0.229072",
+          "temperature": "1",
+          "provider": "Anthropic"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 0.107,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 1.0,
+          "max_attempts": 1
+        }
+      }
+    }
+  ]
+}

flat/objects/00/0c/000caf0d-d4a2-4fd4-b51a-2ea07290822e.json

@@ -0,0 +1,178 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "AC-EVAL/GPT-4/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Beijing University of Posts and Telecommunications",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "GPT-4",
+    "name": "GPT-4",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "AC-EVAL",
+      "source_data": {
+        "dataset_name": "AC-EVAL",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2403.06574"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy of large language models on the AC-EVAL benchmark in a zero-shot, answer-only (AO) setting. This setting evaluates the models' inherent understanding of ancient Chinese without providing any examples in the prompt.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Average Accuracy (%) - Zero-shot AO",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "AC-EVAL: Average Accuracy (Zero-shot, Answer-Only)"
+        },
+        "metric_id": "ac_eval_average_accuracy_zero_shot_answer_only",
+        "metric_name": "AC-EVAL: Average Accuracy (Zero-shot, Answer-Only)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 56.2
+      },
+      "evaluation_result_id": "AC-EVAL/GPT-4/1771591481.616601#ac_eval#ac_eval_average_accuracy_zero_shot_answer_only"
+    },
+    {
+      "evaluation_name": "AC-EVAL",
+      "source_data": {
+        "dataset_name": "AC-EVAL",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2403.06574"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy of large language models on the AC-EVAL benchmark in a five-shot, answer-only (AO) setting. This setting evaluates the models' ability to leverage in-context examples for understanding ancient Chinese. The paper notes that few-shot learning often had a limited or negative impact for this specific benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Average Accuracy (%) - Five-shot AO",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AC-EVAL: Average Accuracy (Five-shot, Answer-Only)"
+        },
+        "metric_id": "ac_eval_average_accuracy_five_shot_answer_only",
+        "metric_name": "AC-EVAL: Average Accuracy (Five-shot, Answer-Only)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 57.45
+      },
+      "evaluation_result_id": "AC-EVAL/GPT-4/1771591481.616601#ac_eval#ac_eval_average_accuracy_five_shot_answer_only"
+    },
+    {
+      "evaluation_name": "AC-EVAL",
+      "source_data": {
+        "dataset_name": "AC-EVAL",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2403.06574"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the 'General Historical Knowledge' category of AC-EVAL, considered the 'easy' difficulty level. This category assesses foundational knowledge of ancient Chinese history, geography, and culture. Evaluated in a zero-shot, answer-only (AO) setting.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - Historical Knowledge (Zero-shot AO)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AC-EVAL: General Historical Knowledge Accuracy (Zero-shot, AO)"
+        },
+        "metric_id": "ac_eval_general_historical_knowledge_accuracy_zero_shot_ao",
+        "metric_name": "AC-EVAL: General Historical Knowledge Accuracy (Zero-shot, AO)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 66.11
+      },
+      "evaluation_result_id": "AC-EVAL/GPT-4/1771591481.616601#ac_eval#ac_eval_general_historical_knowledge_accuracy_zero_shot_ao"
+    },
+    {
+      "evaluation_name": "AC-EVAL",
+      "source_data": {
+        "dataset_name": "AC-EVAL",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2403.06574"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the 'Long Text Comprehension' category of AC-EVAL, considered the 'hard' difficulty level. This category assesses summarization, analysis, and appreciation of classical prose and poetry. This is the most challenging category for all models, evaluated in a zero-shot, answer-only (AO) setting.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - Long Text (Zero-shot AO)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AC-EVAL: Long Text Comprehension Accuracy (Zero-shot, AO)"
+        },
+        "metric_id": "ac_eval_long_text_comprehension_accuracy_zero_shot_ao",
+        "metric_name": "AC-EVAL: Long Text Comprehension Accuracy (Zero-shot, AO)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 47.38
+      },
+      "evaluation_result_id": "AC-EVAL/GPT-4/1771591481.616601#ac_eval#ac_eval_long_text_comprehension_accuracy_zero_shot_ao"
+    },
+    {
+      "evaluation_name": "AC-EVAL",
+      "source_data": {
+        "dataset_name": "AC-EVAL",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2403.06574"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the 'Short Text Understanding' category of AC-EVAL, considered the 'normal' difficulty level. This category assesses lexical semantics, pragmatics, allusions, and translation from short ancient Chinese texts. Evaluated in a zero-shot, answer-only (AO) setting.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - Short Text (Zero-shot AO)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AC-EVAL: Short Text Understanding Accuracy (Zero-shot, AO)"
+        },
+        "metric_id": "ac_eval_short_text_understanding_accuracy_zero_shot_ao",
+        "metric_name": "AC-EVAL: Short Text Understanding Accuracy (Zero-shot, AO)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 55.11
+      },
+      "evaluation_result_id": "AC-EVAL/GPT-4/1771591481.616601#ac_eval#ac_eval_short_text_understanding_accuracy_zero_shot_ao"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/0f/000f114b-6dde-4018-8fd5-650946a6fe90.json

@@ -0,0 +1,208 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "AnaloBench/GPT4-turbo/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Johns Hopkins University",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "GPT4-turbo",
+    "name": "GPT4-turbo",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "AnaloBench",
+      "source_data": {
+        "dataset_name": "AnaloBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12370"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures Mean Average Precision (MAP) for Task 2, where models retrieve the top 10 analogous stories from a large bank of 200. This evaluation uses 10-sentence stories, showing a significant drop in retrieval performance compared to shorter stories.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Mean Average Precision (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AnaloBench: T2 Long-Context Retrieval MAP (10-sentence stories)"
+        },
+        "metric_id": "analobench_t2_long_context_retrieval_map_10_sentence_stories",
+        "metric_name": "AnaloBench: T2 Long-Context Retrieval MAP (10-sentence stories)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 14.2
+      },
+      "evaluation_result_id": "AnaloBench/GPT4-turbo/1771591481.616601#analobench#analobench_t2_long_context_retrieval_map_10_sentence_stories"
+    },
+    {
+      "evaluation_name": "AnaloBench",
+      "source_data": {
+        "dataset_name": "AnaloBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12370"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures Mean Average Precision (MAP) for Task 2, where models retrieve the top 10 analogous stories from a large bank of 200. This evaluation uses 1-sentence stories and assesses the overall quality of the ranked retrieval list.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Mean Average Precision (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AnaloBench: T2 Long-Context Retrieval MAP (1-sentence stories)"
+        },
+        "metric_id": "analobench_t2_long_context_retrieval_map_1_sentence_stories",
+        "metric_name": "AnaloBench: T2 Long-Context Retrieval MAP (1-sentence stories)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 55.4
+      },
+      "evaluation_result_id": "AnaloBench/GPT4-turbo/1771591481.616601#analobench#analobench_t2_long_context_retrieval_map_1_sentence_stories"
+    },
+    {
+      "evaluation_name": "AnaloBench",
+      "source_data": {
+        "dataset_name": "AnaloBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12370"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures Mean Average Precision (MAP) for Task 2, where models retrieve the top 10 analogous stories from a large bank of 200. With 30-sentence stories, model performance is close to the random baseline, highlighting severe limitations in long-context retrieval.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Mean Average Precision (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AnaloBench: T2 Long-Context Retrieval MAP (30-sentence stories)"
+        },
+        "metric_id": "analobench_t2_long_context_retrieval_map_30_sentence_stories",
+        "metric_name": "AnaloBench: T2 Long-Context Retrieval MAP (30-sentence stories)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 10.8
+      },
+      "evaluation_result_id": "AnaloBench/GPT4-turbo/1771591481.616601#analobench#analobench_t2_long_context_retrieval_map_30_sentence_stories"
+    },
+    {
+      "evaluation_name": "AnaloBench",
+      "source_data": {
+        "dataset_name": "AnaloBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12370"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures Mean Reciprocal Rank (MRR) for Task 2, where models retrieve the top 10 analogous stories from a large bank of 200. MRR focuses on the rank of the first correct answer. This evaluation uses 10-sentence stories.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Mean Reciprocal Rank (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AnaloBench: T2 Long-Context Retrieval MRR (10-sentence stories)"
+        },
+        "metric_id": "analobench_t2_long_context_retrieval_mrr_10_sentence_stories",
+        "metric_name": "AnaloBench: T2 Long-Context Retrieval MRR (10-sentence stories)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 15.6
+      },
+      "evaluation_result_id": "AnaloBench/GPT4-turbo/1771591481.616601#analobench#analobench_t2_long_context_retrieval_mrr_10_sentence_stories"
+    },
+    {
+      "evaluation_name": "AnaloBench",
+      "source_data": {
+        "dataset_name": "AnaloBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12370"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures Mean Reciprocal Rank (MRR) for Task 2, where models retrieve the top 10 analogous stories from a large bank of 200. MRR focuses on the rank of the first correct answer. This evaluation uses 1-sentence stories.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Mean Reciprocal Rank (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AnaloBench: T2 Long-Context Retrieval MRR (1-sentence stories)"
+        },
+        "metric_id": "analobench_t2_long_context_retrieval_mrr_1_sentence_stories",
+        "metric_name": "AnaloBench: T2 Long-Context Retrieval MRR (1-sentence stories)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 64.2
+      },
+      "evaluation_result_id": "AnaloBench/GPT4-turbo/1771591481.616601#analobench#analobench_t2_long_context_retrieval_mrr_1_sentence_stories"
+    },
+    {
+      "evaluation_name": "AnaloBench",
+      "source_data": {
+        "dataset_name": "AnaloBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12370"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures Mean Reciprocal Rank (MRR) for Task 2, where models retrieve the top 10 analogous stories from a large bank of 200. MRR focuses on the rank of the first correct answer. With 30-sentence stories, even top models perform near the random baseline.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Mean Reciprocal Rank (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AnaloBench: T2 Long-Context Retrieval MRR (30-sentence stories)"
+        },
+        "metric_id": "analobench_t2_long_context_retrieval_mrr_30_sentence_stories",
+        "metric_name": "AnaloBench: T2 Long-Context Retrieval MRR (30-sentence stories)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 11.3
+      },
+      "evaluation_result_id": "AnaloBench/GPT4-turbo/1771591481.616601#analobench#analobench_t2_long_context_retrieval_mrr_30_sentence_stories"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/11/0011669a-ccce-4fea-9c49-1f29fc5e287e.json

@@ -0,0 +1,238 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "VideoRewardBench/Skywork-VL Reward (Qwen2.5-VL-7B)/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "University of Science and Technology of China",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Skywork-VL Reward (Qwen2.5-VL-7B)",
+    "name": "Skywork-VL Reward (Qwen2.5-VL-7B)",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "VideoRewardBench",
+      "source_data": {
+        "dataset_name": "VideoRewardBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.00484"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall accuracy of 28 multimodal reward models (MRMs) on the VideoRewardBench. This metric measures the proportion of samples where the model's judgment aligns with the human-annotated preference across all 1,563 samples in the benchmark, providing a comprehensive measure of performance.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Accuracy (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Overall Accuracy on VideoRewardBench"
+        },
+        "metric_id": "overall_accuracy_on_videorewardbench",
+        "metric_name": "Overall Accuracy on VideoRewardBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 60.5
+      },
+      "evaluation_result_id": "VideoRewardBench/Skywork-VL Reward (Qwen2.5-VL-7B)/1771591481.616601#videorewardbench#overall_accuracy_on_videorewardbench"
+    },
+    {
+      "evaluation_name": "VideoRewardBench",
+      "source_data": {
+        "dataset_name": "VideoRewardBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.00484"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the Long-Form Perception dimension of VideoRewardBench, assessing the ability to evaluate responses to open-ended, long-form video captioning and understanding tasks. Based on 283 samples.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Long-Form Perception Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Long-Form Perception Accuracy on VideoRewardBench"
+        },
+        "metric_id": "long_form_perception_accuracy_on_videorewardbench",
+        "metric_name": "Long-Form Perception Accuracy on VideoRewardBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 65.7
+      },
+      "evaluation_result_id": "VideoRewardBench/Skywork-VL Reward (Qwen2.5-VL-7B)/1771591481.616601#videorewardbench#long_form_perception_accuracy_on_videorewardbench"
+    },
+    {
+      "evaluation_name": "VideoRewardBench",
+      "source_data": {
+        "dataset_name": "VideoRewardBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.00484"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Macro average accuracy of multimodal reward models (MRMs) on VideoRewardBench, calculated as the arithmetic mean of the accuracies across the five main dimensions (long-form perception, short-form perception, knowledge, reasoning, and safety). This provides a balanced view of performance across different capabilities.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Macro Average Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Macro Average Accuracy on VideoRewardBench"
+        },
+        "metric_id": "macro_average_accuracy_on_videorewardbench",
+        "metric_name": "Macro Average Accuracy on VideoRewardBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 60.4
+      },
+      "evaluation_result_id": "VideoRewardBench/Skywork-VL Reward (Qwen2.5-VL-7B)/1771591481.616601#videorewardbench#macro_average_accuracy_on_videorewardbench"
+    },
+    {
+      "evaluation_name": "VideoRewardBench",
+      "source_data": {
+        "dataset_name": "VideoRewardBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.00484"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the Knowledge dimension of VideoRewardBench, which evaluates tasks requiring specialized domain knowledge. Responses often involve Chain-of-Thought reasoning. Based on 238 samples.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Knowledge Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Knowledge Accuracy on VideoRewardBench"
+        },
+        "metric_id": "knowledge_accuracy_on_videorewardbench",
+        "metric_name": "Knowledge Accuracy on VideoRewardBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 52.9
+      },
+      "evaluation_result_id": "VideoRewardBench/Skywork-VL Reward (Qwen2.5-VL-7B)/1771591481.616601#videorewardbench#knowledge_accuracy_on_videorewardbench"
+    },
+    {
+      "evaluation_name": "VideoRewardBench",
+      "source_data": {
+        "dataset_name": "VideoRewardBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.00484"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the Reasoning dimension of VideoRewardBench, which evaluates abilities across spatial, temporal, relation, and object reasoning. Based on 278 samples.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Reasoning Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Reasoning Accuracy on VideoRewardBench"
+        },
+        "metric_id": "reasoning_accuracy_on_videorewardbench",
+        "metric_name": "Reasoning Accuracy on VideoRewardBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 54
+      },
+      "evaluation_result_id": "VideoRewardBench/Skywork-VL Reward (Qwen2.5-VL-7B)/1771591481.616601#videorewardbench#reasoning_accuracy_on_videorewardbench"
+    },
+    {
+      "evaluation_name": "VideoRewardBench",
+      "source_data": {
+        "dataset_name": "VideoRewardBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.00484"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the Safety dimension of VideoRewardBench, which assesses the awareness of safety issues in video-text multimodal attack scenarios. Based on 351 samples.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Safety Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Safety Accuracy on VideoRewardBench"
+        },
+        "metric_id": "safety_accuracy_on_videorewardbench",
+        "metric_name": "Safety Accuracy on VideoRewardBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 80.1
+      },
+      "evaluation_result_id": "VideoRewardBench/Skywork-VL Reward (Qwen2.5-VL-7B)/1771591481.616601#videorewardbench#safety_accuracy_on_videorewardbench"
+    },
+    {
+      "evaluation_name": "VideoRewardBench",
+      "source_data": {
+        "dataset_name": "VideoRewardBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.00484"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the Short-Form Perception dimension of VideoRewardBench, evaluating responses to short-form video QA tasks like spatiotemporal perception and hallucination detection. Based on 413 samples.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Short-Form Perception Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Short-Form Perception Accuracy on VideoRewardBench"
+        },
+        "metric_id": "short_form_perception_accuracy_on_videorewardbench",
+        "metric_name": "Short-Form Perception Accuracy on VideoRewardBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 49.2
+      },
+      "evaluation_result_id": "VideoRewardBench/Skywork-VL Reward (Qwen2.5-VL-7B)/1771591481.616601#videorewardbench#short_form_perception_accuracy_on_videorewardbench"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/14/00147c87-da76-4476-ba59-58f4e84c27d8.json

@@ -0,0 +1,88 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "SMART-101/Swin-T + Emb./1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Mitsubishi Electric Research Labs",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Swin-T + Emb.",
+    "name": "Swin-T + Emb.",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "SMART-101",
+      "source_data": {
+        "dataset_name": "SMART-101",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2212.09993"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Option selection accuracy (O_acc) of various vision-language models on the Instance Split of SMART-101. This split represents a standard supervised learning setting where models are trained and tested on different instances of the same 101 puzzles. Higher is better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Option Selection Accuracy (O_acc %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Option Selection Accuracy on SMART-101 (Instance Split)"
+        },
+        "metric_id": "option_selection_accuracy_on_smart_101_instance_split",
+        "metric_name": "Option Selection Accuracy on SMART-101 (Instance Split)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 30.8
+      },
+      "evaluation_result_id": "SMART-101/Swin-T + Emb./1771591481.616601#smart_101#option_selection_accuracy_on_smart_101_instance_split"
+    },
+    {
+      "evaluation_name": "SMART-101",
+      "source_data": {
+        "dataset_name": "SMART-101",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2212.09993"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Exact solution accuracy (S_acc) of various vision-language models on the Instance Split of SMART-101. This split represents a standard supervised learning setting. S_acc measures the frequency of producing the exact correct numerical or symbolic solution. Higher is better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Solution Accuracy (S_acc %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Solution Accuracy on SMART-101 (Instance Split)"
+        },
+        "metric_id": "solution_accuracy_on_smart_101_instance_split",
+        "metric_name": "Solution Accuracy on SMART-101 (Instance Split)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 22.5
+      },
+      "evaluation_result_id": "SMART-101/Swin-T + Emb./1771591481.616601#smart_101#solution_accuracy_on_smart_101_instance_split"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/1b/001bbae3-9f38-4bc7-bad6-3812f52664b2.json

@@ -0,0 +1,208 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "FACT-BENCH/Falcon-180B-chat/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "AWS AI Labs",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Falcon-180B-chat",
+    "name": "Falcon-180B-chat",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "FACT-BENCH",
+      "source_data": {
+        "dataset_name": "FACT-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/wikipedia"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Exact Match (EM) scores on the FACT-BENCH PREMIUM2K subset in a 10-shot setting. This metric measures the percentage of model predictions that perfectly match the ground-truth answer when provided with 10 in-context examples.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Exact Match (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Factual Recall on PREMIUM2K (10-shot EM)"
+        },
+        "metric_id": "factual_recall_on_premium2k_10_shot_em",
+        "metric_name": "Factual Recall on PREMIUM2K (10-shot EM)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 49.3
+      },
+      "evaluation_result_id": "FACT-BENCH/Falcon-180B-chat/1771591481.616601#fact_bench#factual_recall_on_premium2k_10_shot_em"
+    },
+    {
+      "evaluation_name": "FACT-BENCH",
+      "source_data": {
+        "dataset_name": "FACT-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/wikipedia"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Contains scores on the FACT-BENCH PREMIUM2K subset in a zero-shot setting. This metric checks if the ground-truth answer (or its aliases) is present anywhere in the model's generated output, which is useful for evaluating verbose models.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Contains Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Factual Recall on PREMIUM2K (0-shot Contains)"
+        },
+        "metric_id": "factual_recall_on_premium2k_0_shot_contains",
+        "metric_name": "Factual Recall on PREMIUM2K (0-shot Contains)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 47.1
+      },
+      "evaluation_result_id": "FACT-BENCH/Falcon-180B-chat/1771591481.616601#fact_bench#factual_recall_on_premium2k_0_shot_contains"
+    },
+    {
+      "evaluation_name": "FACT-BENCH",
+      "source_data": {
+        "dataset_name": "FACT-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/wikipedia"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Contains scores on the FACT-BENCH PREMIUM2K subset in a 10-shot setting. This metric checks if the ground-truth answer is present in the model's output when given 10 in-context examples. It is useful for evaluating verbose models.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Contains Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Factual Recall on PREMIUM2K (10-shot Contains)"
+        },
+        "metric_id": "factual_recall_on_premium2k_10_shot_contains",
+        "metric_name": "Factual Recall on PREMIUM2K (10-shot Contains)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 51.6
+      },
+      "evaluation_result_id": "FACT-BENCH/Falcon-180B-chat/1771591481.616601#fact_bench#factual_recall_on_premium2k_10_shot_contains"
+    },
+    {
+      "evaluation_name": "FACT-BENCH",
+      "source_data": {
+        "dataset_name": "FACT-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/wikipedia"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Exact Match (EM) scores on the FACT-BENCH PREMIUM2K subset in a zero-shot setting. This metric measures the percentage of model predictions that perfectly match the ground-truth answer without any in-context examples provided.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Exact Match (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Factual Recall on PREMIUM2K (0-shot EM)"
+        },
+        "metric_id": "factual_recall_on_premium2k_0_shot_em",
+        "metric_name": "Factual Recall on PREMIUM2K (0-shot EM)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 39.95
+      },
+      "evaluation_result_id": "FACT-BENCH/Falcon-180B-chat/1771591481.616601#fact_bench#factual_recall_on_premium2k_0_shot_em"
+    },
+    {
+      "evaluation_name": "FACT-BENCH",
+      "source_data": {
+        "dataset_name": "FACT-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/wikipedia"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "F1 scores on the FACT-BENCH PREMIUM2K subset in a 10-shot setting. This metric measures the harmonic mean of precision and recall when the model is provided with 10 in-context examples to guide its output.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1 Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Factual Recall on PREMIUM2K (10-shot F1)"
+        },
+        "metric_id": "factual_recall_on_premium2k_10_shot_f1",
+        "metric_name": "Factual Recall on PREMIUM2K (10-shot F1)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 56.17
+      },
+      "evaluation_result_id": "FACT-BENCH/Falcon-180B-chat/1771591481.616601#fact_bench#factual_recall_on_premium2k_10_shot_f1"
+    },
+    {
+      "evaluation_name": "FACT-BENCH",
+      "source_data": {
+        "dataset_name": "FACT-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/wikipedia"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "F1 scores on the FACT-BENCH PREMIUM2K subset in a zero-shot setting. This metric measures the harmonic mean of precision and recall, accounting for partial matches between the model's prediction and the ground-truth answer.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1 Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Factual Recall on PREMIUM2K (0-shot F1)"
+        },
+        "metric_id": "factual_recall_on_premium2k_0_shot_f1",
+        "metric_name": "Factual Recall on PREMIUM2K (0-shot F1)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 48.14
+      },
+      "evaluation_result_id": "FACT-BENCH/Falcon-180B-chat/1771591481.616601#fact_bench#factual_recall_on_premium2k_0_shot_f1"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/20/002008db-63d6-4a75-b039-6d0e1f625e20.json

@@ -0,0 +1,148 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "COESOT/ToMP101/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Chinese Academy of Sciences",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "ToMP101",
+    "name": "ToMP101",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "COESOT",
+      "source_data": {
+        "dataset_name": "COESOT",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2211.11010"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The BreakOut Capability (BOC) score is a novel metric proposed with the COESOT benchmark. It assigns a higher weight to challenging videos where baseline trackers struggle, thus better reflecting an algorithm's outstanding ability compared to baselines. A higher score is better. This evaluation is performed on the COESOT test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "BreakOut Capability (BOC) Score",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "BreakOut Capability (BOC) Score on the COESOT Benchmark"
+        },
+        "metric_id": "breakout_capability_boc_score_on_the_coesot_benchmark",
+        "metric_name": "BreakOut Capability (BOC) Score on the COESOT Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 18.3
+      },
+      "evaluation_result_id": "COESOT/ToMP101/1771591481.616601#coesot#breakout_capability_boc_score_on_the_coesot_benchmark"
+    },
+    {
+      "evaluation_name": "COESOT",
+      "source_data": {
+        "dataset_name": "COESOT",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2211.11010"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Normalized Precision Rate (NPR) normalizes the precision rate to mitigate dependency on target size and image resolution, ensuring consistency across different scales. A higher score is better. This evaluation is performed on the COESOT test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Normalized Precision Rate (NPR)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Normalized Precision Rate (NPR) on the COESOT Benchmark"
+        },
+        "metric_id": "normalized_precision_rate_npr_on_the_coesot_benchmark",
+        "metric_name": "Normalized Precision Rate (NPR) on the COESOT Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 71.3
+      },
+      "evaluation_result_id": "COESOT/ToMP101/1771591481.616601#coesot#normalized_precision_rate_npr_on_the_coesot_benchmark"
+    },
+    {
+      "evaluation_name": "COESOT",
+      "source_data": {
+        "dataset_name": "COESOT",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2211.11010"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Precision Rate (PR) is the percentage of frames where the center location error between the predicted and ground truth bounding box is below a predefined threshold (20 pixels). A higher score indicates better performance. This evaluation is performed on the COESOT test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Precision Rate (PR)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Precision Rate (PR) on the COESOT Benchmark"
+        },
+        "metric_id": "precision_rate_pr_on_the_coesot_benchmark",
+        "metric_name": "Precision Rate (PR) on the COESOT Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 71.6
+      },
+      "evaluation_result_id": "COESOT/ToMP101/1771591481.616601#coesot#precision_rate_pr_on_the_coesot_benchmark"
+    },
+    {
+      "evaluation_name": "COESOT",
+      "source_data": {
+        "dataset_name": "COESOT",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2211.11010"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success Rate (SR) measures the average overlap ratio (IoU) between the predicted and ground truth bounding boxes. A higher score indicates better performance. This evaluation is performed on the test set of COESOT, a large-scale benchmark for color-event based single object tracking.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (SR)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Success Rate (SR) on the COESOT Benchmark"
+        },
+        "metric_id": "success_rate_sr_on_the_coesot_benchmark",
+        "metric_name": "Success Rate (SR) on the COESOT Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 59.9
+      },
+      "evaluation_result_id": "COESOT/ToMP101/1771591481.616601#coesot#success_rate_sr_on_the_coesot_benchmark"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/20/0020dba1-4ff9-4da5-8e99-b79f2b481c70.json

@@ -0,0 +1,931 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407",
+  "retrieved_timestamp": "1775236112.40407",
+  "source_metadata": {
+    "source_name": "BFCL leaderboard CSV",
+    "source_type": "documentation",
+    "source_organization_name": "UC Berkeley Gorilla",
+    "source_organization_url": "https://gorilla.cs.berkeley.edu/leaderboard.html",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "csv_url": "https://gorilla.cs.berkeley.edu/data_overall.csv",
+      "leaderboard_url": "https://gorilla.cs.berkeley.edu/leaderboard.html",
+      "leaderboard_version": "BFCL V4"
+    }
+  },
+  "eval_library": {
+    "name": "BFCL",
+    "version": "v4"
+  },
+  "model_info": {
+    "name": "Gemini-2.5-Flash-Lite (Prompt)",
+    "id": "google/gemini-2-5-flash-lite-prompt",
+    "developer": "google",
+    "additional_details": {
+      "raw_model_name": "Gemini-2.5-Flash-Lite (Prompt)",
+      "organization": "Google",
+      "license": "Proprietary",
+      "mode": "Prompt",
+      "model_link": "https://deepmind.google/technologies/gemini/flash-lite/"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#overall#bfcl_overall_rank",
+      "evaluation_name": "overall",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.overall.rank",
+        "metric_name": "Overall rank",
+        "metric_kind": "rank",
+        "metric_unit": "position",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 1.0,
+        "max_score": 109.0,
+        "additional_details": {
+          "raw_metric_field": "Rank",
+          "raw_evaluation_name": "bfcl.overall.rank"
+        }
+      },
+      "score_details": {
+        "score": 73.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#overall#bfcl_overall_overall_accuracy",
+      "evaluation_name": "overall",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.overall.overall_accuracy",
+        "metric_name": "Overall accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Overall Acc",
+          "raw_evaluation_name": "bfcl.overall.overall_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 28.03
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#overall#bfcl_overall_total_cost_usd",
+      "evaluation_name": "overall",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.overall.total_cost_usd",
+        "metric_name": "Total cost",
+        "metric_kind": "cost",
+        "metric_unit": "usd",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 355.17,
+        "additional_details": {
+          "raw_metric_field": "Total Cost ($)",
+          "raw_evaluation_name": "bfcl.overall.total_cost_usd"
+        }
+      },
+      "score_details": {
+        "score": 7.05
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#overall#bfcl_overall_latency_mean_s",
+      "evaluation_name": "overall",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.overall.latency_mean_s",
+        "metric_name": "Latency mean",
+        "metric_kind": "latency",
+        "metric_unit": "seconds",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 169.87,
+        "additional_details": {
+          "raw_metric_field": "Latency Mean (s)",
+          "raw_evaluation_name": "bfcl.overall.latency_mean_s"
+        }
+      },
+      "score_details": {
+        "score": 1.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#overall#bfcl_overall_latency_std_s",
+      "evaluation_name": "overall",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.overall.latency_std_s",
+        "metric_name": "Latency standard deviation",
+        "metric_kind": "latency",
+        "metric_unit": "seconds",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 212.99,
+        "additional_details": {
+          "raw_metric_field": "Latency Standard Deviation (s)",
+          "raw_evaluation_name": "bfcl.overall.latency_std_s"
+        }
+      },
+      "score_details": {
+        "score": 4.75
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#overall#bfcl_overall_latency_p95_s",
+      "evaluation_name": "overall",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.overall.latency_p95_s",
+        "metric_name": "Latency 95th percentile",
+        "metric_kind": "latency",
+        "metric_unit": "seconds",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 568.59,
+        "additional_details": {
+          "raw_metric_field": "Latency 95th Percentile (s)",
+          "raw_evaluation_name": "bfcl.overall.latency_p95_s"
+        }
+      },
+      "score_details": {
+        "score": 1.4
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#non_live#bfcl_non_live_ast_accuracy",
+      "evaluation_name": "non_live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.non_live.ast_accuracy",
+        "metric_name": "Non-live AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Non-Live AST Acc",
+          "raw_evaluation_name": "bfcl.non_live.ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 83.9
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#non_live#bfcl_non_live_simple_ast_accuracy",
+      "evaluation_name": "non_live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.non_live.simple_ast_accuracy",
+        "metric_name": "Non-live simple AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Non-Live Simple AST",
+          "raw_evaluation_name": "bfcl.non_live.simple_ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 70.08
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#non_live#bfcl_non_live_multiple_ast_accuracy",
+      "evaluation_name": "non_live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.non_live.multiple_ast_accuracy",
+        "metric_name": "Non-live multiple AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Non-Live Multiple AST",
+          "raw_evaluation_name": "bfcl.non_live.multiple_ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 86.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#non_live#bfcl_non_live_parallel_ast_accuracy",
+      "evaluation_name": "non_live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.non_live.parallel_ast_accuracy",
+        "metric_name": "Non-live parallel AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Non-Live Parallel AST",
+          "raw_evaluation_name": "bfcl.non_live.parallel_ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 90.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#non_live#bfcl_non_live_parallel_multiple_ast_accuracy",
+      "evaluation_name": "non_live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.non_live.parallel_multiple_ast_accuracy",
+        "metric_name": "Non-live parallel multiple AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Non-Live Parallel Multiple AST",
+          "raw_evaluation_name": "bfcl.non_live.parallel_multiple_ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 89.5
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#live#bfcl_live_live_accuracy",
+      "evaluation_name": "live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.live.live_accuracy",
+        "metric_name": "Live accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Live Acc",
+          "raw_evaluation_name": "bfcl.live.live_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 54.85
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#live#bfcl_live_live_simple_ast_accuracy",
+      "evaluation_name": "live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.live.live_simple_ast_accuracy",
+        "metric_name": "Live simple AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Live Simple AST",
+          "raw_evaluation_name": "bfcl.live.live_simple_ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 67.05
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#live#bfcl_live_live_multiple_ast_accuracy",
+      "evaluation_name": "live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.live.live_multiple_ast_accuracy",
+        "metric_name": "Live multiple AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Live Multiple AST",
+          "raw_evaluation_name": "bfcl.live.live_multiple_ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 51.66
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#live#bfcl_live_live_parallel_ast_accuracy",
+      "evaluation_name": "live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.live.live_parallel_ast_accuracy",
+        "metric_name": "Live parallel AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Live Parallel AST",
+          "raw_evaluation_name": "bfcl.live.live_parallel_ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 75.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#live#bfcl_live_live_parallel_multiple_ast_accuracy",
+      "evaluation_name": "live",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.live.live_parallel_multiple_ast_accuracy",
+        "metric_name": "Live parallel multiple AST accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Live Parallel Multiple AST",
+          "raw_evaluation_name": "bfcl.live.live_parallel_multiple_ast_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 50.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#multi_turn#bfcl_multi_turn_accuracy",
+      "evaluation_name": "multi_turn",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.multi_turn.accuracy",
+        "metric_name": "Multi-turn accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Multi Turn Acc",
+          "raw_evaluation_name": "bfcl.multi_turn.accuracy"
+        }
+      },
+      "score_details": {
+        "score": 7.63
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#multi_turn#bfcl_multi_turn_base_accuracy",
+      "evaluation_name": "multi_turn",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.multi_turn.base_accuracy",
+        "metric_name": "Multi-turn base accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Multi Turn Base",
+          "raw_evaluation_name": "bfcl.multi_turn.base_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 10.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#multi_turn#bfcl_multi_turn_miss_function_accuracy",
+      "evaluation_name": "multi_turn",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.multi_turn.miss_function_accuracy",
+        "metric_name": "Multi-turn missing function accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Multi Turn Miss Func",
+          "raw_evaluation_name": "bfcl.multi_turn.miss_function_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 5.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#multi_turn#bfcl_multi_turn_miss_parameter_accuracy",
+      "evaluation_name": "multi_turn",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.multi_turn.miss_parameter_accuracy",
+        "metric_name": "Multi-turn missing parameter accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Multi Turn Miss Param",
+          "raw_evaluation_name": "bfcl.multi_turn.miss_parameter_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 6.5
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#multi_turn#bfcl_multi_turn_long_context_accuracy",
+      "evaluation_name": "multi_turn",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.multi_turn.long_context_accuracy",
+        "metric_name": "Multi-turn long-context accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Multi Turn Long Context",
+          "raw_evaluation_name": "bfcl.multi_turn.long_context_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 9.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#web_search#bfcl_web_search_accuracy",
+      "evaluation_name": "web_search",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.web_search.accuracy",
+        "metric_name": "Web-search accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Web Search Acc",
+          "raw_evaluation_name": "bfcl.web_search.accuracy"
+        }
+      },
+      "score_details": {
+        "score": 0.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#web_search#bfcl_web_search_base_accuracy",
+      "evaluation_name": "web_search",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.web_search.base_accuracy",
+        "metric_name": "Web-search base accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Web Search Base",
+          "raw_evaluation_name": "bfcl.web_search.base_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 0.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#web_search#bfcl_web_search_no_snippet_accuracy",
+      "evaluation_name": "web_search",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.web_search.no_snippet_accuracy",
+        "metric_name": "Web-search no-snippet accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Web Search No Snippet",
+          "raw_evaluation_name": "bfcl.web_search.no_snippet_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 0.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#memory#bfcl_memory_accuracy",
+      "evaluation_name": "memory",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.memory.accuracy",
+        "metric_name": "Memory accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Memory Acc",
+          "raw_evaluation_name": "bfcl.memory.accuracy"
+        }
+      },
+      "score_details": {
+        "score": 12.69
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#memory#bfcl_memory_kv_accuracy",
+      "evaluation_name": "memory",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.memory.kv_accuracy",
+        "metric_name": "Memory KV accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Memory KV",
+          "raw_evaluation_name": "bfcl.memory.kv_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 1.94
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#memory#bfcl_memory_vector_accuracy",
+      "evaluation_name": "memory",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.memory.vector_accuracy",
+        "metric_name": "Memory vector accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Memory Vector",
+          "raw_evaluation_name": "bfcl.memory.vector_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 6.45
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#memory#bfcl_memory_recursive_summarization_accuracy",
+      "evaluation_name": "memory",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.memory.recursive_summarization_accuracy",
+        "metric_name": "Memory recursive summarization accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Memory Recursive Summarization",
+          "raw_evaluation_name": "bfcl.memory.recursive_summarization_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 29.68
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#relevance#bfcl_relevance_relevance_detection_accuracy",
+      "evaluation_name": "relevance",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.relevance.relevance_detection_accuracy",
+        "metric_name": "Relevance detection accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Relevance Detection",
+          "raw_evaluation_name": "bfcl.relevance.relevance_detection_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 50.0
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#relevance#bfcl_relevance_irrelevance_detection_accuracy",
+      "evaluation_name": "relevance",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.relevance.irrelevance_detection_accuracy",
+        "metric_name": "Irrelevance detection accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percentage",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Irrelevance Detection",
+          "raw_evaluation_name": "bfcl.relevance.irrelevance_detection_accuracy"
+        }
+      },
+      "score_details": {
+        "score": 93.33
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#format_sensitivity#bfcl_format_sensitivity_max_delta",
+      "evaluation_name": "format_sensitivity",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.format_sensitivity.max_delta",
+        "metric_name": "Format sensitivity max delta",
+        "metric_kind": "difference",
+        "metric_unit": "percentage_points",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Format Sensitivity Max Delta",
+          "raw_evaluation_name": "bfcl.format_sensitivity.max_delta"
+        }
+      },
+      "score_details": {
+        "score": 25.5
+      }
+    },
+    {
+      "evaluation_result_id": "bfcl/google/gemini-2-5-flash-lite-prompt/1775236112.40407#format_sensitivity#bfcl_format_sensitivity_stddev",
+      "evaluation_name": "format_sensitivity",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "BFCL leaderboard CSV",
+        "url": [
+          "https://gorilla.cs.berkeley.edu/data_overall.csv"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "bfcl.format_sensitivity.stddev",
+        "metric_name": "Format sensitivity standard deviation",
+        "metric_kind": "difference",
+        "metric_unit": "percentage_points",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "raw_metric_field": "Format Sensitivity Standard Deviation",
+          "raw_evaluation_name": "bfcl.format_sensitivity.stddev"
+        }
+      },
+      "score_details": {
+        "score": 6.68
+      }
+    }
+  ]
+}

flat/objects/00/21/002127a3-a6ec-40a2-9227-7a25ac276f92.json

@@ -0,0 +1,148 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "ThaiOCRBench/InternVL3 14B/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "SCB 10X",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "InternVL3 14B",
+    "name": "InternVL3 14B",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "ThaiOCRBench",
+      "source_data": {
+        "dataset_name": "ThaiOCRBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2511.04479"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average Tree Edit Distance (TED) score across structural understanding tasks (Chart parsing, Table parsing, Document parsing) from the ThaiOCRBench benchmark. TED measures the structural similarity between predicted and reference outputs for hierarchical content like tables and documents. Higher scores indicate better performance.",
+        "additional_details": {
+          "alphaxiv_y_axis": "TED Score (Avg)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Structural Understanding Performance on ThaiOCRBench (TED)"
+        },
+        "metric_id": "structural_understanding_performance_on_thaiocrbench_ted",
+        "metric_name": "Structural Understanding Performance on ThaiOCRBench (TED)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.76
+      },
+      "evaluation_result_id": "ThaiOCRBench/InternVL3 14B/1771591481.616601#thaiocrbench#structural_understanding_performance_on_thaiocrbench_ted"
+    },
+    {
+      "evaluation_name": "ThaiOCRBench",
+      "source_data": {
+        "dataset_name": "ThaiOCRBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2511.04479"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average F1-score across information extraction tasks (Key information extraction, Key information mapping) from the ThaiOCRBench benchmark. This metric evaluates the precision and recall of entity-level predictions, which is crucial for tasks requiring exact field alignment. Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1-score (Avg)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Information Extraction Performance on ThaiOCRBench (F1)"
+        },
+        "metric_id": "information_extraction_performance_on_thaiocrbench_f1",
+        "metric_name": "Information Extraction Performance on ThaiOCRBench (F1)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.405
+      },
+      "evaluation_result_id": "ThaiOCRBench/InternVL3 14B/1771591481.616601#thaiocrbench#information_extraction_performance_on_thaiocrbench_f1"
+    },
+    {
+      "evaluation_name": "ThaiOCRBench",
+      "source_data": {
+        "dataset_name": "ThaiOCRBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2511.04479"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average BMFL score across text generation and recognition tasks (Fine-grained Recognition, Full-page OCR, Handwritten, Text recognition) from the ThaiOCRBench benchmark. BMFL is a composite metric averaging BLEU, METEOR, F1-score, and Normalized Levenshtein Similarity, assessing character-level accuracy and linguistic fidelity. Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "BMFL Score (Avg)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Text Generation and Recognition on ThaiOCRBench (BMFL)"
+        },
+        "metric_id": "text_generation_and_recognition_on_thaiocrbench_bmfl",
+        "metric_name": "Text Generation and Recognition on ThaiOCRBench (BMFL)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.399
+      },
+      "evaluation_result_id": "ThaiOCRBench/InternVL3 14B/1771591481.616601#thaiocrbench#text_generation_and_recognition_on_thaiocrbench_bmfl"
+    },
+    {
+      "evaluation_name": "ThaiOCRBench",
+      "source_data": {
+        "dataset_name": "ThaiOCRBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2511.04479"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average Normalized Levenshtein Similarity (ANLS) score across understanding and VQA tasks (Document classification, Diagram VQA, Cognition VQA, Infographics VQA) from the ThaiOCRBench benchmark. ANLS measures the similarity between predicted and reference text responses, allowing for partial credit. Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "ANLS Score (Avg)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Visual Question Answering Performance on ThaiOCRBench (ANLS)"
+        },
+        "metric_id": "visual_question_answering_performance_on_thaiocrbench_anls",
+        "metric_name": "Visual Question Answering Performance on ThaiOCRBench (ANLS)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.059
+      },
+      "evaluation_result_id": "ThaiOCRBench/InternVL3 14B/1771591481.616601#thaiocrbench#visual_question_answering_performance_on_thaiocrbench_anls"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/27/0027aeb0-f093-4a15-82ff-2b17bd29d42e.json

@@ -0,0 +1,222 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "vals-ai/tax_eval_v2/anthropic_claude-opus-4-1-20250805-thinking/1777395187.3170502",
+  "retrieved_timestamp": "1777395187.3170502",
+  "source_metadata": {
+    "source_name": "Vals.ai Leaderboard - TaxEval (v2)",
+    "source_type": "documentation",
+    "source_organization_name": "Vals.ai",
+    "source_organization_url": "https://www.vals.ai",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "benchmark_slug": "tax_eval_v2",
+      "benchmark_name": "TaxEval (v2)",
+      "benchmark_updated": "2026-04-16",
+      "dataset_type": "private",
+      "industry": "finance",
+      "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2",
+      "extraction_method": "static_astro_benchmark_view_props"
+    }
+  },
+  "eval_library": {
+    "name": "Vals.ai",
+    "version": "unknown"
+  },
+  "model_info": {
+    "name": "claude-opus-4-1-20250805-thinking",
+    "id": "anthropic/claude-opus-4-1-20250805-thinking",
+    "developer": "anthropic",
+    "additional_details": {
+      "vals_model_id": "anthropic/claude-opus-4-1-20250805-thinking",
+      "vals_provider": "Anthropic"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_result_id": "tax_eval_v2:correctness:anthropic/claude-opus-4-1-20250805-thinking:score",
+      "evaluation_name": "vals_ai.tax_eval_v2.correctness",
+      "source_data": {
+        "dataset_name": "TaxEval (v2) - Correctness",
+        "source_type": "other",
+        "additional_details": {
+          "benchmark_slug": "tax_eval_v2",
+          "task_key": "correctness",
+          "dataset_type": "private",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for TaxEval (v2) (Correctness).",
+        "metric_id": "vals_ai.tax_eval_v2.correctness.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2"
+        }
+      },
+      "score_details": {
+        "score": 62.96,
+        "details": {
+          "benchmark_slug": "tax_eval_v2",
+          "benchmark_name": "TaxEval (v2)",
+          "benchmark_updated": "2026-04-16",
+          "task_key": "correctness",
+          "task_name": "Correctness",
+          "dataset_type": "private",
+          "industry": "finance",
+          "raw_score": "62.96",
+          "raw_stderr": "1.381",
+          "latency": "26.637",
+          "cost_per_test": "0.093482",
+          "temperature": "1",
+          "max_output_tokens": "32000",
+          "provider": "Anthropic"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.381,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 1.0,
+          "max_tokens": 32000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "tax_eval_v2:overall:anthropic/claude-opus-4-1-20250805-thinking:score",
+      "evaluation_name": "vals_ai.tax_eval_v2.overall",
+      "source_data": {
+        "dataset_name": "TaxEval (v2) - Overall",
+        "source_type": "other",
+        "additional_details": {
+          "benchmark_slug": "tax_eval_v2",
+          "task_key": "overall",
+          "dataset_type": "private",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for TaxEval (v2) (Overall).",
+        "metric_id": "vals_ai.tax_eval_v2.overall.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2"
+        }
+      },
+      "score_details": {
+        "score": 73.671,
+        "details": {
+          "benchmark_slug": "tax_eval_v2",
+          "benchmark_name": "TaxEval (v2)",
+          "benchmark_updated": "2026-04-16",
+          "task_key": "overall",
+          "task_name": "Overall",
+          "dataset_type": "private",
+          "industry": "finance",
+          "raw_score": "73.671",
+          "raw_stderr": "0.864",
+          "latency": "26.637",
+          "cost_per_test": "0.093482",
+          "temperature": "1",
+          "max_output_tokens": "32000",
+          "provider": "Anthropic"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 0.864,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 1.0,
+          "max_tokens": 32000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "tax_eval_v2:stepwise_reasoning:anthropic/claude-opus-4-1-20250805-thinking:score",
+      "evaluation_name": "vals_ai.tax_eval_v2.stepwise_reasoning",
+      "source_data": {
+        "dataset_name": "TaxEval (v2) - Stepwise Reasoning",
+        "source_type": "other",
+        "additional_details": {
+          "benchmark_slug": "tax_eval_v2",
+          "task_key": "stepwise_reasoning",
+          "dataset_type": "private",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for TaxEval (v2) (Stepwise Reasoning).",
+        "metric_id": "vals_ai.tax_eval_v2.stepwise_reasoning.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/tax_eval_v2"
+        }
+      },
+      "score_details": {
+        "score": 84.383,
+        "details": {
+          "benchmark_slug": "tax_eval_v2",
+          "benchmark_name": "TaxEval (v2)",
+          "benchmark_updated": "2026-04-16",
+          "task_key": "stepwise_reasoning",
+          "task_name": "Stepwise Reasoning",
+          "dataset_type": "private",
+          "industry": "finance",
+          "raw_score": "84.383",
+          "raw_stderr": "1.038",
+          "latency": "26.637",
+          "cost_per_test": "0.093482",
+          "temperature": "1",
+          "max_output_tokens": "32000",
+          "provider": "Anthropic"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.038,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 1.0,
+          "max_tokens": 32000,
+          "max_attempts": 1
+        }
+      }
+    }
+  ]
+}

flat/objects/00/28/002873a0-91ed-4892-bbbd-fe5d28ef686b.json

@@ -0,0 +1,58 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "OlympiadBench/imagination-research/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Beihang University",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "imagination-research",
+    "name": "imagination-research",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "OlympiadBench",
+      "source_data": {
+        "dataset_name": "OlympiadBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.14008"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Final scores of the top five teams on the private leaderboard of the AIMO Progress Prize 2 competition. The evaluation is based on the number of correctly solved problems out of a set of 50 Olympiad-level math problems.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Score (out of 50) - Private Leaderboard",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "AIMO Progress Prize 2 - Private Leaderboard"
+        },
+        "metric_id": "aimo_progress_prize_2_private_leaderboard",
+        "metric_name": "AIMO Progress Prize 2 - Private Leaderboard",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 31
+      },
+      "evaluation_result_id": "OlympiadBench/imagination-research/1771591481.616601#olympiadbench#aimo_progress_prize_2_private_leaderboard"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/2c/002c7439-2c20-473e-938b-bf0421cccc71.json

@@ -0,0 +1,58 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "Alvorada-Bench/Claude 3.5 Sonnet/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Inteli",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Claude 3.5 Sonnet",
+    "name": "Claude 3.5 Sonnet",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "Alvorada-Bench",
+      "source_data": {
+        "dataset_name": "Alvorada-Bench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HenriqueGodoy/Alvorada-bench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Mean accuracy of language models on the Alvorada-Bench, a dataset of 4,515 multiple-choice questions from five major Brazilian university entrance exams (ENEM, FUVEST, UNICAMP, IME, ITA). The evaluation assesses factual knowledge, reasoning, and cultural understanding in Brazilian Portuguese.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Alvorada-Bench Overall Accuracy on Brazilian University Entrance Exams"
+        },
+        "metric_id": "alvorada_bench_overall_accuracy_on_brazilian_university_entrance_exams",
+        "metric_name": "Alvorada-Bench Overall Accuracy on Brazilian University Entrance Exams",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.7941
+      },
+      "evaluation_result_id": "Alvorada-Bench/Claude 3.5 Sonnet/1771591481.616601#alvorada_bench#alvorada_bench_overall_accuracy_on_brazilian_university_entrance_exams"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/2c/002c98e5-dd11-4602-8f52-c383fe8b61b9.json

@@ -0,0 +1,88 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "ESGenius/Llama-3.2 (3B)/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Alibaba Group",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Llama-3.2 (3B)",
+    "name": "Llama-3.2 (3B)",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "ESGenius",
+      "source_data": {
+        "dataset_name": "ESGenius",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.01646"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the accuracy of large language models on the ESGenius benchmark without any provided context (zero-shot). This evaluates the models' intrinsic knowledge of Environmental, Social, and Governance (ESG) topics. The benchmark consists of 1,136 expert-validated multiple-choice questions. Accuracy is reported as the percentage of questions answered correctly.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "ESGenius: Zero-Shot Question Answering Accuracy"
+        },
+        "metric_id": "esgenius_zero_shot_question_answering_accuracy",
+        "metric_name": "ESGenius: Zero-Shot Question Answering Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 60.74
+      },
+      "evaluation_result_id": "ESGenius/Llama-3.2 (3B)/1771591481.616601#esgenius#esgenius_zero_shot_question_answering_accuracy"
+    },
+    {
+      "evaluation_name": "ESGenius",
+      "source_data": {
+        "dataset_name": "ESGenius",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.01646"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the accuracy of large language models on the ESGenius benchmark when provided with relevant context documents (Retrieval-Augmented Generation). This evaluates the models' ability to synthesize information from provided authoritative sources to answer ESG-related questions. The benchmark consists of 1,136 expert-validated multiple-choice questions. Accuracy is reported as the percentage of questions answered correctly.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ESGenius: RAG Question Answering Accuracy"
+        },
+        "metric_id": "esgenius_rag_question_answering_accuracy",
+        "metric_name": "ESGenius: RAG Question Answering Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 68.31
+      },
+      "evaluation_result_id": "ESGenius/Llama-3.2 (3B)/1771591481.616601#esgenius#esgenius_rag_question_answering_accuracy"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/2d/002dbe87-27c5-4dca-9657-a54b713aad37.json

@@ -0,0 +1,118 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "BearCubs/GPT-4o + Google Search/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "University of Maryland, College Park",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "GPT-4o + Google Search",
+    "name": "GPT-4o + Google Search",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "BearCubs",
+      "source_data": {
+        "dataset_name": "BearCubs",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2503.07919"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall accuracy on the full BEARCUBS benchmark, which consists of 111 information-seeking questions (56 text-based, 55 multimodal) requiring live web interaction. This metric represents the percentage of questions for which the agent provided a correct, unambiguous answer.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - All Questions",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "BEARCUBS: Overall Accuracy on All Questions"
+        },
+        "metric_id": "bearcubs_overall_accuracy_on_all_questions",
+        "metric_name": "BEARCUBS: Overall Accuracy on All Questions",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0
+      },
+      "evaluation_result_id": "BearCubs/GPT-4o + Google Search/1771591481.616601#bearcubs#bearcubs_overall_accuracy_on_all_questions"
+    },
+    {
+      "evaluation_name": "BearCubs",
+      "source_data": {
+        "dataset_name": "BearCubs",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2503.07919"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the 55 multimodal questions from the BEARCUBS benchmark. These tasks require agents to interpret various media formats like images, videos, audio, or interactive elements (e.g., games, virtual tours) that cannot be solved via text-based workarounds.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - Multimodal Questions",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "BEARCUBS: Accuracy on Multimodal Questions"
+        },
+        "metric_id": "bearcubs_accuracy_on_multimodal_questions",
+        "metric_name": "BEARCUBS: Accuracy on Multimodal Questions",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0
+      },
+      "evaluation_result_id": "BearCubs/GPT-4o + Google Search/1771591481.616601#bearcubs#bearcubs_accuracy_on_multimodal_questions"
+    },
+    {
+      "evaluation_name": "BearCubs",
+      "source_data": {
+        "dataset_name": "BearCubs",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2503.07919"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the 56 text-based questions from the BEARCUBS benchmark. These tasks involve reading and navigating text-heavy web content, such as online databases or articles, to find a factual answer.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - Text Only Questions",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "BEARCUBS: Accuracy on Text-Only Questions"
+        },
+        "metric_id": "bearcubs_accuracy_on_text_only_questions",
+        "metric_name": "BEARCUBS: Accuracy on Text-Only Questions",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0
+      },
+      "evaluation_result_id": "BearCubs/GPT-4o + Google Search/1771591481.616601#bearcubs#bearcubs_accuracy_on_text_only_questions"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/2e/002e9fc6-7ed2-434b-9e4d-8951a40c4578.json

@@ -0,0 +1,748 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "RULEARN/Llama3-70B/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "University of Texas at Dallas",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Llama3-70B",
+    "name": "Llama3-70B",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall success rate across all three puzzle types on the RULEARN benchmark using the proposed IDEA agent. The IDEA framework integrates induction, deduction, and abduction to improve the rule-learning ability of LLMs. This is the primary result demonstrating the paper's contribution.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - All Types - IDEA Agent",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Overall Success Rate on RULEARN Benchmark (IDEA Agent)"
+        },
+        "metric_id": "overall_success_rate_on_rulearn_benchmark_idea_agent",
+        "metric_name": "Overall Success Rate on RULEARN Benchmark (IDEA Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 29
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#overall_success_rate_on_rulearn_benchmark_idea_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle across all RULEARN puzzles for the proposed IDEA agent. Lower scores indicate more efficient exploration. This metric shows IDEA reduces redundant actions.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - All Puzzles - IDEA Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on All RULEARN Puzzles (IDEA Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_all_rulearn_puzzles_idea_agent",
+        "metric_name": "Average Repeated Actions on All RULEARN Puzzles (IDEA Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.73
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_all_rulearn_puzzles_idea_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle across all RULEARN puzzles for the ReAct agent (Baseline). Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - All Puzzles - ReAct Agent (Baseline)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on All RULEARN Puzzles (ReAct Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_all_rulearn_puzzles_react_agent",
+        "metric_name": "Average Repeated Actions on All RULEARN Puzzles (ReAct Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 3.36
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_all_rulearn_puzzles_react_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Escape Room puzzles for the Deduction Only agent. Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Escape Room - Deduction Only Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Escape Room Puzzles (Deduction Only Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_escape_room_puzzles_deduction_only_agent",
+        "metric_name": "Average Repeated Actions on Escape Room Puzzles (Deduction Only Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.44
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_escape_room_puzzles_deduction_only_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Escape Room puzzles for the proposed IDEA agent. Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Escape Room - IDEA Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Escape Room Puzzles (IDEA Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_escape_room_puzzles_idea_agent",
+        "metric_name": "Average Repeated Actions on Escape Room Puzzles (IDEA Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.72
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_escape_room_puzzles_idea_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Escape Room puzzles for the ReAct agent (Baseline). Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Escape Room - ReAct Agent (Baseline)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Escape Room Puzzles (ReAct Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_escape_room_puzzles_react_agent",
+        "metric_name": "Average Repeated Actions on Escape Room Puzzles (ReAct Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.59
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_escape_room_puzzles_react_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Function Operator puzzles for the Deduction Only agent. Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Function Operator - Deduction Only Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Function Operator Puzzles (Deduction Only Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_function_operator_puzzles_deduction_only_agent",
+        "metric_name": "Average Repeated Actions on Function Operator Puzzles (Deduction Only Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 2.47
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_function_operator_puzzles_deduction_only_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Function Operator puzzles for the proposed IDEA agent. Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Function Operator - IDEA Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Function Operator Puzzles (IDEA Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_function_operator_puzzles_idea_agent",
+        "metric_name": "Average Repeated Actions on Function Operator Puzzles (IDEA Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.43
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_function_operator_puzzles_idea_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Function Operator puzzles for the ReAct agent (Baseline). Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Function Operator - ReAct Agent (Baseline)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Function Operator Puzzles (ReAct Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_function_operator_puzzles_react_agent",
+        "metric_name": "Average Repeated Actions on Function Operator Puzzles (ReAct Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.25
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_function_operator_puzzles_react_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Reactor puzzles for the Deduction Only agent. Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Reactor - Deduction Only Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Reactor Puzzles (Deduction Only Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_reactor_puzzles_deduction_only_agent",
+        "metric_name": "Average Repeated Actions on Reactor Puzzles (Deduction Only Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 6.25
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_reactor_puzzles_deduction_only_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Reactor puzzles for the proposed IDEA agent. Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Reactor - IDEA Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Reactor Puzzles (IDEA Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_reactor_puzzles_idea_agent",
+        "metric_name": "Average Repeated Actions on Reactor Puzzles (IDEA Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 4.05
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_reactor_puzzles_idea_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle on the Reactor puzzles for the ReAct agent (Baseline). Lower scores indicate more efficient exploration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - Reactor - ReAct Agent (Baseline)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on Reactor Puzzles (ReAct Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_reactor_puzzles_react_agent",
+        "metric_name": "Average Repeated Actions on Reactor Puzzles (ReAct Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 7.23
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_reactor_puzzles_react_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall success rate across all three puzzle types (Function Operator, Escape Room, Reactor) on the RULEARN benchmark. This evaluation uses the Oracle-rule agent, where the ground-truth rule is explicitly provided to the LLM. This measures the agent's ability to apply a known rule, not learn it.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - All Types - Oracle-rule Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Success Rate on RULEARN Benchmark (Oracle-rule Agent)"
+        },
+        "metric_id": "overall_success_rate_on_rulearn_benchmark_oracle_rule_agent",
+        "metric_name": "Overall Success Rate on RULEARN Benchmark (Oracle-rule Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 32.67
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#overall_success_rate_on_rulearn_benchmark_oracle_rule_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall success rate across all three puzzle types (Function Operator, Escape Room, Reactor) on the RULEARN benchmark. This evaluation uses the ReAct agent as a baseline, which must learn the rules from interaction without an explicit reasoning framework like IDEA.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - All Types - ReAct Agent (Baseline)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Success Rate on RULEARN Benchmark (ReAct Agent)"
+        },
+        "metric_id": "overall_success_rate_on_rulearn_benchmark_react_agent",
+        "metric_name": "Overall Success Rate on RULEARN Benchmark (ReAct Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 19.67
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#overall_success_rate_on_rulearn_benchmark_react_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Escape Room puzzles of the RULEARN benchmark using the proposed IDEA agent. The IDEA framework integrates induction, deduction, and abduction to improve the rule-learning ability of LLMs.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Escape Room - IDEA Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Escape Room Success Rate on RULEARN Benchmark (IDEA Agent)"
+        },
+        "metric_id": "escape_room_success_rate_on_rulearn_benchmark_idea_agent",
+        "metric_name": "Escape Room Success Rate on RULEARN Benchmark (IDEA Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 35
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#escape_room_success_rate_on_rulearn_benchmark_idea_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Escape Room puzzles of the RULEARN benchmark. This evaluation uses the Oracle-rule agent, where the ground-truth rule (how the password is constructed) is explicitly provided to the LLM. This measures the agent's ability to apply a known rule to find the password.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Escape Room - Oracle-rule Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Escape Room Success Rate on RULEARN Benchmark (Oracle-rule Agent)"
+        },
+        "metric_id": "escape_room_success_rate_on_rulearn_benchmark_oracle_rule_agent",
+        "metric_name": "Escape Room Success Rate on RULEARN Benchmark (Oracle-rule Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 48
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#escape_room_success_rate_on_rulearn_benchmark_oracle_rule_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Escape Room puzzles of the RULEARN benchmark. This evaluation uses the ReAct agent as a baseline, which must learn the rules from interaction without an explicit reasoning framework like IDEA.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Escape Room - ReAct Agent (Baseline)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Escape Room Success Rate on RULEARN Benchmark (ReAct Agent)"
+        },
+        "metric_id": "escape_room_success_rate_on_rulearn_benchmark_react_agent",
+        "metric_name": "Escape Room Success Rate on RULEARN Benchmark (ReAct Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 17
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#escape_room_success_rate_on_rulearn_benchmark_react_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Function Operator puzzles of the RULEARN benchmark using the proposed IDEA agent. The IDEA framework integrates induction, deduction, and abduction to improve the rule-learning ability of LLMs.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Function Operator - IDEA Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Function Operator Success Rate on RULEARN Benchmark (IDEA Agent)"
+        },
+        "metric_id": "function_operator_success_rate_on_rulearn_benchmark_idea_agent",
+        "metric_name": "Function Operator Success Rate on RULEARN Benchmark (IDEA Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 41
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#function_operator_success_rate_on_rulearn_benchmark_idea_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Function Operator puzzles of the RULEARN benchmark. This evaluation uses the Oracle-rule agent, where the ground-truth rule (function form) is explicitly provided to the LLM. This measures the agent's ability to apply a known rule to deduce coefficients.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Function Operator - Oracle-rule Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Function Operator Success Rate on RULEARN Benchmark (Oracle-rule Agent)"
+        },
+        "metric_id": "function_operator_success_rate_on_rulearn_benchmark_oracle_rule_agent",
+        "metric_name": "Function Operator Success Rate on RULEARN Benchmark (Oracle-rule Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 33
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#function_operator_success_rate_on_rulearn_benchmark_oracle_rule_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Function Operator puzzles of the RULEARN benchmark. This evaluation uses the ReAct agent as a baseline, which must learn the rules from interaction without an explicit reasoning framework like IDEA.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Function Operator - ReAct Agent (Baseline)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Function Operator Success Rate on RULEARN Benchmark (ReAct Agent)"
+        },
+        "metric_id": "function_operator_success_rate_on_rulearn_benchmark_react_agent",
+        "metric_name": "Function Operator Success Rate on RULEARN Benchmark (ReAct Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 33
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#function_operator_success_rate_on_rulearn_benchmark_react_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Reactor puzzles of the RULEARN benchmark using the proposed IDEA agent. The IDEA framework integrates induction, deduction, and abduction to improve the rule-learning ability of LLMs.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Reactor - IDEA Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Reactor Success Rate on RULEARN Benchmark (IDEA Agent)"
+        },
+        "metric_id": "reactor_success_rate_on_rulearn_benchmark_idea_agent",
+        "metric_name": "Reactor Success Rate on RULEARN Benchmark (IDEA Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 11
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#reactor_success_rate_on_rulearn_benchmark_idea_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Reactor puzzles of the RULEARN benchmark. This evaluation uses the Oracle-rule agent, where the ground-truth rule (string combination logic) is explicitly provided to the LLM. This measures the agent's ability to apply a known rule to synthesize a target string.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Reactor - Oracle-rule Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Reactor Success Rate on RULEARN Benchmark (Oracle-rule Agent)"
+        },
+        "metric_id": "reactor_success_rate_on_rulearn_benchmark_oracle_rule_agent",
+        "metric_name": "Reactor Success Rate on RULEARN Benchmark (Oracle-rule Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 17
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#reactor_success_rate_on_rulearn_benchmark_oracle_rule_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The average number of repeated actions per puzzle across all RULEARN puzzles for the Deduction Only agent. Lower scores indicate more efficient exploration. This metric provides insight into agent behavior.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Repeated Actions - All Puzzles - Deduction Only Agent",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Average Repeated Actions on All RULEARN Puzzles (Deduction Only Agent)"
+        },
+        "metric_id": "average_repeated_actions_on_all_rulearn_puzzles_deduction_only_agent",
+        "metric_name": "Average Repeated Actions on All RULEARN Puzzles (Deduction Only Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 3.39
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#average_repeated_actions_on_all_rulearn_puzzles_deduction_only_agent"
+    },
+    {
+      "evaluation_name": "RULEARN",
+      "source_data": {
+        "dataset_name": "RULEARN",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2408.10455"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Success rate on the Reactor puzzles of the RULEARN benchmark. This evaluation uses the ReAct agent as a baseline, which must learn the rules from interaction without an explicit reasoning framework like IDEA.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Success Rate (%) - Reactor - ReAct Agent (Baseline)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Reactor Success Rate on RULEARN Benchmark (ReAct Agent)"
+        },
+        "metric_id": "reactor_success_rate_on_rulearn_benchmark_react_agent",
+        "metric_name": "Reactor Success Rate on RULEARN Benchmark (ReAct Agent)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 9
+      },
+      "evaluation_result_id": "RULEARN/Llama3-70B/1771591481.616601#rulearn#reactor_success_rate_on_rulearn_benchmark_react_agent"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/34/00341beb-c613-42d5-91fe-f6b392ff4ea7.json

@@ -0,0 +1,118 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "Video-MMLU/XComposer2 (InternLM2-7B)/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "University of Washington",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "XComposer2 (InternLM2-7B)",
+    "name": "XComposer2 (InternLM2-7B)",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "Video-MMLU",
+      "source_data": {
+        "dataset_name": "Video-MMLU",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Enxin/Video-MMLU"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The overall performance score on the Video-MMLU benchmark, calculated as the average of the 'Notebook' (detailed captioning) and 'Quiz' (reasoning QA) scores. This metric provides a balanced assessment of a model's ability to both perceive fine-grained visual details and perform complex reasoning on lecture videos. Scores are on a 0-100 scale, derived from an underlying 0-5 LLM-based evaluation.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Score",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Video-MMLU: Overall Performance"
+        },
+        "metric_id": "video_mmlu_overall_performance",
+        "metric_name": "Video-MMLU: Overall Performance",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 25.62
+      },
+      "evaluation_result_id": "Video-MMLU/XComposer2 (InternLM2-7B)/1771591481.616601#video_mmlu#video_mmlu_overall_performance"
+    },
+    {
+      "evaluation_name": "Video-MMLU",
+      "source_data": {
+        "dataset_name": "Video-MMLU",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Enxin/Video-MMLU"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the 'Review Notes' (detailed captioning) task of the Video-MMLU benchmark. This metric, known as VDCscore, evaluates a model's ability to generate detailed, accurate descriptions of lecture videos, focusing on visual perception of elements like formulas, text, and dynamic demonstrations. The score is an average from an LLM-based evaluator on a 0-100 scale, with strict rules for OCR accuracy.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Notebook Average Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Video-MMLU: Detailed Captioning Performance (Notebook Avg)"
+        },
+        "metric_id": "video_mmlu_detailed_captioning_performance_notebook_avg",
+        "metric_name": "Video-MMLU: Detailed Captioning Performance (Notebook Avg)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 16.24
+      },
+      "evaluation_result_id": "Video-MMLU/XComposer2 (InternLM2-7B)/1771591481.616601#video_mmlu#video_mmlu_detailed_captioning_performance_notebook_avg"
+    },
+    {
+      "evaluation_name": "Video-MMLU",
+      "source_data": {
+        "dataset_name": "Video-MMLU",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Enxin/Video-MMLU"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the 'Take Quiz' (reasoning QA) task of the Video-MMLU benchmark. This metric evaluates a model's ability to answer complex, open-ended questions that require deep understanding and reasoning about the concepts presented in lecture videos, going beyond surface-level visual features. The score is an average from an LLM-based evaluator on a 0-100 scale, where answers must capture all critical concepts to be marked correct.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Quiz Average Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Video-MMLU: Reasoning Question Answering Performance (Quiz Avg)"
+        },
+        "metric_id": "video_mmlu_reasoning_question_answering_performance_quiz_avg",
+        "metric_name": "Video-MMLU: Reasoning Question Answering Performance (Quiz Avg)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 35
+      },
+      "evaluation_result_id": "Video-MMLU/XComposer2 (InternLM2-7B)/1771591481.616601#video_mmlu#video_mmlu_reasoning_question_answering_performance_quiz_avg"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/35/0035179a-05c1-4403-9fae-b274fedc9712.json

@@ -0,0 +1,118 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "BearCubs/Google Deep Research/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "University of Maryland, College Park",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Google Deep Research",
+    "name": "Google Deep Research",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "BearCubs",
+      "source_data": {
+        "dataset_name": "BearCubs",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2503.07919"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall accuracy on the full BEARCUBS benchmark, which consists of 111 information-seeking questions (56 text-based, 55 multimodal) requiring live web interaction. This metric represents the percentage of questions for which the agent provided a correct, unambiguous answer.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - All Questions",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "BEARCUBS: Overall Accuracy on All Questions"
+        },
+        "metric_id": "bearcubs_overall_accuracy_on_all_questions",
+        "metric_name": "BEARCUBS: Overall Accuracy on All Questions",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 23.4
+      },
+      "evaluation_result_id": "BearCubs/Google Deep Research/1771591481.616601#bearcubs#bearcubs_overall_accuracy_on_all_questions"
+    },
+    {
+      "evaluation_name": "BearCubs",
+      "source_data": {
+        "dataset_name": "BearCubs",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2503.07919"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the 55 multimodal questions from the BEARCUBS benchmark. These tasks require agents to interpret various media formats like images, videos, audio, or interactive elements (e.g., games, virtual tours) that cannot be solved via text-based workarounds.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - Multimodal Questions",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "BEARCUBS: Accuracy on Multimodal Questions"
+        },
+        "metric_id": "bearcubs_accuracy_on_multimodal_questions",
+        "metric_name": "BEARCUBS: Accuracy on Multimodal Questions",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 3.6
+      },
+      "evaluation_result_id": "BearCubs/Google Deep Research/1771591481.616601#bearcubs#bearcubs_accuracy_on_multimodal_questions"
+    },
+    {
+      "evaluation_name": "BearCubs",
+      "source_data": {
+        "dataset_name": "BearCubs",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2503.07919"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the 56 text-based questions from the BEARCUBS benchmark. These tasks involve reading and navigating text-heavy web content, such as online databases or articles, to find a factual answer.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - Text Only Questions",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "BEARCUBS: Accuracy on Text-Only Questions"
+        },
+        "metric_id": "bearcubs_accuracy_on_text_only_questions",
+        "metric_name": "BEARCUBS: Accuracy on Text-Only Questions",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 42.9
+      },
+      "evaluation_result_id": "BearCubs/Google Deep Research/1771591481.616601#bearcubs#bearcubs_accuracy_on_text_only_questions"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/36/0036f037-d1c4-4a6b-8dca-c37a5fc5e58b.json

@@ -0,0 +1,169 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "hfopenllm_v2/microsoft_Phi-3-small-128k-instruct/1773936498.240187",
+  "retrieved_timestamp": "1773936498.240187",
+  "source_metadata": {
+    "source_name": "HF Open LLM v2",
+    "source_type": "documentation",
+    "source_organization_name": "Hugging Face",
+    "evaluator_relationship": "third_party"
+  },
+  "eval_library": {
+    "name": "lm-evaluation-harness",
+    "version": "0.4.0",
+    "additional_details": {
+      "fork": "https://github.com/huggingface/lm-evaluation-harness/tree/adding_all_changess"
+    }
+  },
+  "model_info": {
+    "name": "Phi-3-small-128k-instruct",
+    "id": "microsoft/Phi-3-small-128k-instruct",
+    "developer": "microsoft",
+    "inference_platform": "unknown",
+    "additional_details": {
+      "precision": "bfloat16",
+      "architecture": "Phi3SmallForCausalLM",
+      "params_billions": "7.392"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "IFEval",
+      "source_data": {
+        "dataset_name": "IFEval",
+        "source_type": "hf_dataset",
+        "hf_repo": "google/IFEval"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on IFEval",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.6368
+      },
+      "evaluation_result_id": "hfopenllm_v2/microsoft_Phi-3-small-128k-instruct/1773936498.240187#ifeval#accuracy"
+    },
+    {
+      "evaluation_name": "BBH",
+      "source_data": {
+        "dataset_name": "BBH",
+        "source_type": "hf_dataset",
+        "hf_repo": "SaylorTwift/bbh"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on BBH",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.6202
+      },
+      "evaluation_result_id": "hfopenllm_v2/microsoft_Phi-3-small-128k-instruct/1773936498.240187#bbh#accuracy"
+    },
+    {
+      "evaluation_name": "MATH Level 5",
+      "source_data": {
+        "dataset_name": "MATH Level 5",
+        "source_type": "hf_dataset",
+        "hf_repo": "DigitalLearningGmbH/MATH-lighteval"
+      },
+      "metric_config": {
+        "evaluation_description": "Exact Match on MATH Level 5",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "exact_match",
+        "metric_name": "Exact Match",
+        "metric_kind": "exact_match",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.2026
+      },
+      "evaluation_result_id": "hfopenllm_v2/microsoft_Phi-3-small-128k-instruct/1773936498.240187#math_level_5#exact_match"
+    },
+    {
+      "evaluation_name": "GPQA",
+      "source_data": {
+        "dataset_name": "GPQA",
+        "source_type": "hf_dataset",
+        "hf_repo": "Idavidrein/gpqa"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on GPQA",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.3171
+      },
+      "evaluation_result_id": "hfopenllm_v2/microsoft_Phi-3-small-128k-instruct/1773936498.240187#gpqa#accuracy"
+    },
+    {
+      "evaluation_name": "MUSR",
+      "source_data": {
+        "dataset_name": "MUSR",
+        "source_type": "hf_dataset",
+        "hf_repo": "TAUR-Lab/MuSR"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on MUSR",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.4378
+      },
+      "evaluation_result_id": "hfopenllm_v2/microsoft_Phi-3-small-128k-instruct/1773936498.240187#musr#accuracy"
+    },
+    {
+      "evaluation_name": "MMLU-PRO",
+      "source_data": {
+        "dataset_name": "MMLU-PRO",
+        "source_type": "hf_dataset",
+        "hf_repo": "TIGER-Lab/MMLU-Pro"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on MMLU-PRO",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.4491
+      },
+      "evaluation_result_id": "hfopenllm_v2/microsoft_Phi-3-small-128k-instruct/1773936498.240187#mmlu_pro#accuracy"
+    }
+  ]
+}

flat/objects/00/38/00383737-aa12-40d5-8aa4-54d710dbc8e7.json

@@ -0,0 +1,658 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "RECV/GPT-4o/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "The Alan Turing Institute",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "GPT-4o",
+    "name": "GPT-4o",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Error rate on the subset of the CLIMATE-FEVER dataset requiring abductive reasoning, a key failure point identified by the paper. A lower score indicates better performance and stronger reasoning capabilities. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Abductive Error Rate (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Abductive Reasoning Error Rate on the CLIMATE-FEVER Dataset"
+        },
+        "metric_id": "abductive_reasoning_error_rate_on_the_climate_fever_dataset",
+        "metric_name": "Abductive Reasoning Error Rate on the CLIMATE-FEVER Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 25.49
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#abductive_reasoning_error_rate_on_the_climate_fever_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Error rate on the subset of the PHEMEPlus dataset requiring abductive reasoning for claim verification. A lower score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Abductive Error Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Abductive Reasoning Error Rate on the PHEMEPlus Dataset"
+        },
+        "metric_id": "abductive_reasoning_error_rate_on_the_phemeplus_dataset",
+        "metric_name": "Abductive Reasoning Error Rate on the PHEMEPlus Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 41.67
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#abductive_reasoning_error_rate_on_the_phemeplus_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Error rate on the subset of the VitaminC dataset requiring abductive reasoning for claim verification. A lower score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Abductive Error Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Abductive Reasoning Error Rate on the VitaminC Dataset"
+        },
+        "metric_id": "abductive_reasoning_error_rate_on_the_vitaminc_dataset",
+        "metric_name": "Abductive Reasoning Error Rate on the VitaminC Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 30
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#abductive_reasoning_error_rate_on_the_vitaminc_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Estimates the coherence and effectiveness of integrating information from the evidence in LLM-generated rationales on the CLIMATE-FEVER dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "BARTScore",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "BARTScore of Generated Rationales on CLIMATE-FEVER"
+        },
+        "metric_id": "bartscore_of_generated_rationales_on_climate_fever",
+        "metric_name": "BARTScore of Generated Rationales on CLIMATE-FEVER",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": -3.39
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#bartscore_of_generated_rationales_on_climate_fever"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Estimates the coherence and effectiveness of integrating information from the evidence in LLM-generated rationales on the PHEMEPlus dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "BARTScore",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "BARTScore of Generated Rationales on PHEMEPlus"
+        },
+        "metric_id": "bartscore_of_generated_rationales_on_phemeplus",
+        "metric_name": "BARTScore of Generated Rationales on PHEMEPlus",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": -3.74
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#bartscore_of_generated_rationales_on_phemeplus"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Estimates the coherence and effectiveness of integrating information from the evidence in LLM-generated rationales on the VitaminC dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "BARTScore",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "BARTScore of Generated Rationales on VitaminC"
+        },
+        "metric_id": "bartscore_of_generated_rationales_on_vitaminc",
+        "metric_name": "BARTScore of Generated Rationales on VitaminC",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": -3.45
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#bartscore_of_generated_rationales_on_vitaminc"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Error rate on the subset of the CLIMATE-FEVER dataset requiring deductive reasoning for claim verification. A lower score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Deductive Error Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Deductive Reasoning Error Rate on the CLIMATE-FEVER Dataset"
+        },
+        "metric_id": "deductive_reasoning_error_rate_on_the_climate_fever_dataset",
+        "metric_name": "Deductive Reasoning Error Rate on the CLIMATE-FEVER Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 9.05
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#deductive_reasoning_error_rate_on_the_climate_fever_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Error rate on the subset of the PHEMEPlus dataset requiring deductive reasoning for claim verification. A lower score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Deductive Error Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Deductive Reasoning Error Rate on the PHEMEPlus Dataset"
+        },
+        "metric_id": "deductive_reasoning_error_rate_on_the_phemeplus_dataset",
+        "metric_name": "Deductive Reasoning Error Rate on the PHEMEPlus Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 18.32
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#deductive_reasoning_error_rate_on_the_phemeplus_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Error rate on the subset of the VitaminC dataset requiring deductive reasoning for claim verification. A lower score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Deductive Error Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Deductive Reasoning Error Rate on the VitaminC Dataset"
+        },
+        "metric_id": "deductive_reasoning_error_rate_on_the_vitaminc_dataset",
+        "metric_name": "Deductive Reasoning Error Rate on the VitaminC Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 8.94
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#deductive_reasoning_error_rate_on_the_vitaminc_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures how well LLM-generated rationales leverage the provided evidence on the CLIMATE-FEVER dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Evidence Appropriateness (EA)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Evidence Appropriateness of Generated Rationales on CLIMATE-FEVER"
+        },
+        "metric_id": "evidence_appropriateness_of_generated_rationales_on_climate_fever",
+        "metric_name": "Evidence Appropriateness of Generated Rationales on CLIMATE-FEVER",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.93
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#evidence_appropriateness_of_generated_rationales_on_climate_fever"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures how well LLM-generated rationales leverage the provided evidence on the PHEMEPlus dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Evidence Appropriateness (EA)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Evidence Appropriateness of Generated Rationales on PHEMEPlus"
+        },
+        "metric_id": "evidence_appropriateness_of_generated_rationales_on_phemeplus",
+        "metric_name": "Evidence Appropriateness of Generated Rationales on PHEMEPlus",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.89
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#evidence_appropriateness_of_generated_rationales_on_phemeplus"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures how well LLM-generated rationales leverage the provided evidence on the VitaminC dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Evidence Appropriateness (EA)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Evidence Appropriateness of Generated Rationales on VitaminC"
+        },
+        "metric_id": "evidence_appropriateness_of_generated_rationales_on_vitaminc",
+        "metric_name": "Evidence Appropriateness of Generated Rationales on VitaminC",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.91
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#evidence_appropriateness_of_generated_rationales_on_vitaminc"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Macro F1 score for claim verification (Supports vs. Refutes) on the CLIMATE-FEVER dataset from the RECV benchmark. This dataset contains claims related to climate science. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1 Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Claim Verification F1 Score on the CLIMATE-FEVER Dataset"
+        },
+        "metric_id": "claim_verification_f1_score_on_the_climate_fever_dataset",
+        "metric_name": "Claim Verification F1 Score on the CLIMATE-FEVER Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.92
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#claim_verification_f1_score_on_the_climate_fever_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Macro F1 score for claim verification (Supports vs. Refutes) on the PHEMEPlus dataset from the RECV benchmark. This dataset contains rumors from social media. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1 Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Claim Verification F1 Score on the PHEMEPlus Dataset"
+        },
+        "metric_id": "claim_verification_f1_score_on_the_phemeplus_dataset",
+        "metric_name": "Claim Verification F1 Score on the PHEMEPlus Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.75
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#claim_verification_f1_score_on_the_phemeplus_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Macro F1 score for claim verification (Supports vs. Refutes) on the VitaminC dataset from the RECV benchmark. This dataset contains general knowledge claims from Wikipedia. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1 Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Claim Verification F1 Score on the VitaminC Dataset"
+        },
+        "metric_id": "claim_verification_f1_score_on_the_vitaminc_dataset",
+        "metric_name": "Claim Verification F1 Score on the VitaminC Dataset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.89
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#claim_verification_f1_score_on_the_vitaminc_dataset"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Assesses the absence of contradiction between LLM-generated rationales and human-written rationales on the CLIMATE-FEVER dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Factual Consistency (FC)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Factual Consistency of Generated Rationales on CLIMATE-FEVER"
+        },
+        "metric_id": "factual_consistency_of_generated_rationales_on_climate_fever",
+        "metric_name": "Factual Consistency of Generated Rationales on CLIMATE-FEVER",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.94
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#factual_consistency_of_generated_rationales_on_climate_fever"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Assesses the absence of contradiction between LLM-generated rationales and human-written rationales on the PHEMEPlus dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Factual Consistency (FC)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Factual Consistency of Generated Rationales on PHEMEPlus"
+        },
+        "metric_id": "factual_consistency_of_generated_rationales_on_phemeplus",
+        "metric_name": "Factual Consistency of Generated Rationales on PHEMEPlus",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.9
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#factual_consistency_of_generated_rationales_on_phemeplus"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Assesses the absence of contradiction between LLM-generated rationales and human-written rationales on the VitaminC dataset. A higher score indicates better performance. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Factual Consistency (FC)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Factual Consistency of Generated Rationales on VitaminC"
+        },
+        "metric_id": "factual_consistency_of_generated_rationales_on_vitaminc",
+        "metric_name": "Factual Consistency of Generated Rationales on VitaminC",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.89
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#factual_consistency_of_generated_rationales_on_vitaminc"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the fluency of the generated rationales on the CLIMATE-FEVER dataset using GPT-2-XL. A lower perplexity indicates higher fluency. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Perplexity (PPL)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Perplexity of Generated Rationales on CLIMATE-FEVER"
+        },
+        "metric_id": "perplexity_of_generated_rationales_on_climate_fever",
+        "metric_name": "Perplexity of Generated Rationales on CLIMATE-FEVER",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 46.92
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#perplexity_of_generated_rationales_on_climate_fever"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the fluency of the generated rationales on the PHEMEPlus dataset using GPT-2-XL. A lower perplexity indicates higher fluency. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Perplexity (PPL)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Perplexity of Generated Rationales on PHEMEPlus"
+        },
+        "metric_id": "perplexity_of_generated_rationales_on_phemeplus",
+        "metric_name": "Perplexity of Generated Rationales on PHEMEPlus",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 52.85
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#perplexity_of_generated_rationales_on_phemeplus"
+    },
+    {
+      "evaluation_name": "RECV",
+      "source_data": {
+        "dataset_name": "RECV",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.10735"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the fluency of the generated rationales on the VitaminC dataset using GPT-2-XL. A lower perplexity indicates higher fluency. The score represents the best performance for each model across various prompting strategies.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Perplexity (PPL)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Perplexity of Generated Rationales on VitaminC"
+        },
+        "metric_id": "perplexity_of_generated_rationales_on_vitaminc",
+        "metric_name": "Perplexity of Generated Rationales on VitaminC",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 35.82
+      },
+      "evaluation_result_id": "RECV/GPT-4o/1771591481.616601#recv#perplexity_of_generated_rationales_on_vitaminc"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/39/0039055f-58eb-4aa9-a1c7-2832224f611a.json

@@ -0,0 +1,1925 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "piqa/mistral_mistral-large-latest/1742646860.0",
+  "evaluation_timestamp": "1742646860.0",
+  "retrieved_timestamp": "1775627856.831822",
+  "source_metadata": {
+    "source_name": "inspect_ai",
+    "source_type": "evaluation_run",
+    "source_organization_name": "Arcadia Impact",
+    "evaluator_relationship": "third_party"
+  },
+  "eval_library": {
+    "name": "inspect_ai",
+    "version": "inspect_ai:0.3.75"
+  },
+  "model_info": {
+    "name": "mistral/mistral-large-latest",
+    "id": "mistral/mistral-large-latest",
+    "developer": "mistral",
+    "inference_platform": "mistral"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "piqa",
+      "source_data": {
+        "dataset_name": "piqa",
+        "source_type": "hf_dataset",
+        "hf_repo": "piqa",
+        "samples_number": 1838,
+        "sample_ids": [
+          "1",
+          "2",
+          "3",
+          "4",
+          "5",
+          "6",
+          "7",
+          "8",
+          "9",
+          "10",
+          "11",
+          "12",
+          "13",
+          "14",
+          "15",
+          "16",
+          "17",
+          "18",
+          "19",
+          "20",
+          "21",
+          "22",
+          "23",
+          "24",
+          "25",
+          "26",
+          "27",
+          "28",
+          "29",
+          "30",
+          "31",
+          "32",
+          "33",
+          "34",
+          "35",
+          "36",
+          "37",
+          "38",
+          "39",
+          "40",
+          "41",
+          "42",
+          "43",
+          "44",
+          "45",
+          "46",
+          "47",
+          "48",
+          "49",
+          "50",
+          "51",
+          "52",
+          "53",
+          "54",
+          "55",
+          "56",
+          "57",
+          "58",
+          "59",
+          "60",
+          "61",
+          "62",
+          "63",
+          "64",
+          "65",
+          "66",
+          "67",
+          "68",
+          "69",
+          "70",
+          "71",
+          "72",
+          "73",
+          "74",
+          "75",
+          "76",
+          "77",
+          "78",
+          "79",
+          "80",
+          "81",
+          "82",
+          "83",
+          "84",
+          "85",
+          "86",
+          "87",
+          "88",
+          "89",
+          "90",
+          "91",
+          "92",
+          "93",
+          "94",
+          "95",
+          "96",
+          "97",
+          "98",
+          "99",
+          "100",
+          "101",
+          "102",
+          "103",
+          "104",
+          "105",
+          "106",
+          "107",
+          "108",
+          "109",
+          "110",
+          "111",
+          "112",
+          "113",
+          "114",
+          "115",
+          "116",
+          "117",
+          "118",
+          "119",
+          "120",
+          "121",
+          "122",
+          "123",
+          "124",
+          "125",
+          "126",
+          "127",
+          "128",
+          "129",
+          "130",
+          "131",
+          "132",
+          "133",
+          "134",
+          "135",
+          "136",
+          "137",
+          "138",
+          "139",
+          "140",
+          "141",
+          "142",
+          "143",
+          "144",
+          "145",
+          "146",
+          "147",
+          "148",
+          "149",
+          "150",
+          "151",
+          "152",
+          "153",
+          "154",
+          "155",
+          "156",
+          "157",
+          "158",
+          "159",
+          "160",
+          "161",
+          "162",
+          "163",
+          "164",
+          "165",
+          "166",
+          "167",
+          "168",
+          "169",
+          "170",
+          "171",
+          "172",
+          "173",
+          "174",
+          "175",
+          "176",
+          "177",
+          "178",
+          "179",
+          "180",
+          "181",
+          "182",
+          "183",
+          "184",
+          "185",
+          "186",
+          "187",
+          "188",
+          "189",
+          "190",
+          "191",
+          "192",
+          "193",
+          "194",
+          "195",
+          "196",
+          "197",
+          "198",
+          "199",
+          "200",
+          "201",
+          "202",
+          "203",
+          "204",
+          "205",
+          "206",
+          "207",
+          "208",
+          "209",
+          "210",
+          "211",
+          "212",
+          "213",
+          "214",
+          "215",
+          "216",
+          "217",
+          "218",
+          "219",
+          "220",
+          "221",
+          "222",
+          "223",
+          "224",
+          "225",
+          "226",
+          "227",
+          "228",
+          "229",
+          "230",
+          "231",
+          "232",
+          "233",
+          "234",
+          "235",
+          "236",
+          "237",
+          "238",
+          "239",
+          "240",
+          "241",
+          "242",
+          "243",
+          "244",
+          "245",
+          "246",
+          "247",
+          "248",
+          "249",
+          "250",
+          "251",
+          "252",
+          "253",
+          "254",
+          "255",
+          "256",
+          "257",
+          "258",
+          "259",
+          "260",
+          "261",
+          "262",
+          "263",
+          "264",
+          "265",
+          "266",
+          "267",
+          "268",
+          "269",
+          "270",
+          "271",
+          "272",
+          "273",
+          "274",
+          "275",
+          "276",
+          "277",
+          "278",
+          "279",
+          "280",
+          "281",
+          "282",
+          "283",
+          "284",
+          "285",
+          "286",
+          "287",
+          "288",
+          "289",
+          "290",
+          "291",
+          "292",
+          "293",
+          "294",
+          "295",
+          "296",
+          "297",
+          "298",
+          "299",
+          "300",
+          "301",
+          "302",
+          "303",
+          "304",
+          "305",
+          "306",
+          "307",
+          "308",
+          "309",
+          "310",
+          "311",
+          "312",
+          "313",
+          "314",
+          "315",
+          "316",
+          "317",
+          "318",
+          "319",
+          "320",
+          "321",
+          "322",
+          "323",
+          "324",
+          "325",
+          "326",
+          "327",
+          "328",
+          "329",
+          "330",
+          "331",
+          "332",
+          "333",
+          "334",
+          "335",
+          "336",
+          "337",
+          "338",
+          "339",
+          "340",
+          "341",
+          "342",
+          "343",
+          "344",
+          "345",
+          "346",
+          "347",
+          "348",
+          "349",
+          "350",
+          "351",
+          "352",
+          "353",
+          "354",
+          "355",
+          "356",
+          "357",
+          "358",
+          "359",
+          "360",
+          "361",
+          "362",
+          "363",
+          "364",
+          "365",
+          "366",
+          "367",
+          "368",
+          "369",
+          "370",
+          "371",
+          "372",
+          "373",
+          "374",
+          "375",
+          "376",
+          "377",
+          "378",
+          "379",
+          "380",
+          "381",
+          "382",
+          "383",
+          "384",
+          "385",
+          "386",
+          "387",
+          "388",
+          "389",
+          "390",
+          "391",
+          "392",
+          "393",
+          "394",
+          "395",
+          "396",
+          "397",
+          "398",
+          "399",
+          "400",
+          "401",
+          "402",
+          "403",
+          "404",
+          "405",
+          "406",
+          "407",
+          "408",
+          "409",
+          "410",
+          "411",
+          "412",
+          "413",
+          "414",
+          "415",
+          "416",
+          "417",
+          "418",
+          "419",
+          "420",
+          "421",
+          "422",
+          "423",
+          "424",
+          "425",
+          "426",
+          "427",
+          "428",
+          "429",
+          "430",
+          "431",
+          "432",
+          "433",
+          "434",
+          "435",
+          "436",
+          "437",
+          "438",
+          "439",
+          "440",
+          "441",
+          "442",
+          "443",
+          "444",
+          "445",
+          "446",
+          "447",
+          "448",
+          "449",
+          "450",
+          "451",
+          "452",
+          "453",
+          "454",
+          "455",
+          "456",
+          "457",
+          "458",
+          "459",
+          "460",
+          "461",
+          "462",
+          "463",
+          "464",
+          "465",
+          "466",
+          "467",
+          "468",
+          "469",
+          "470",
+          "471",
+          "472",
+          "473",
+          "474",
+          "475",
+          "476",
+          "477",
+          "478",
+          "479",
+          "480",
+          "481",
+          "482",
+          "483",
+          "484",
+          "485",
+          "486",
+          "487",
+          "488",
+          "489",
+          "490",
+          "491",
+          "492",
+          "493",
+          "494",
+          "495",
+          "496",
+          "497",
+          "498",
+          "499",
+          "500",
+          "501",
+          "502",
+          "503",
+          "504",
+          "505",
+          "506",
+          "507",
+          "508",
+          "509",
+          "510",
+          "511",
+          "512",
+          "513",
+          "514",
+          "515",
+          "516",
+          "517",
+          "518",
+          "519",
+          "520",
+          "521",
+          "522",
+          "523",
+          "524",
+          "525",
+          "526",
+          "527",
+          "528",
+          "529",
+          "530",
+          "531",
+          "532",
+          "533",
+          "534",
+          "535",
+          "536",
+          "537",
+          "538",
+          "539",
+          "540",
+          "541",
+          "542",
+          "543",
+          "544",
+          "545",
+          "546",
+          "547",
+          "548",
+          "549",
+          "550",
+          "551",
+          "552",
+          "553",
+          "554",
+          "555",
+          "556",
+          "557",
+          "558",
+          "559",
+          "560",
+          "561",
+          "562",
+          "563",
+          "564",
+          "565",
+          "566",
+          "567",
+          "568",
+          "569",
+          "570",
+          "571",
+          "572",
+          "573",
+          "574",
+          "575",
+          "576",
+          "577",
+          "578",
+          "579",
+          "580",
+          "581",
+          "582",
+          "583",
+          "584",
+          "585",
+          "586",
+          "587",
+          "588",
+          "589",
+          "590",
+          "591",
+          "592",
+          "593",
+          "594",
+          "595",
+          "596",
+          "597",
+          "598",
+          "599",
+          "600",
+          "601",
+          "602",
+          "603",
+          "604",
+          "605",
+          "606",
+          "607",
+          "608",
+          "609",
+          "610",
+          "611",
+          "612",
+          "613",
+          "614",
+          "615",
+          "616",
+          "617",
+          "618",
+          "619",
+          "620",
+          "621",
+          "622",
+          "623",
+          "624",
+          "625",
+          "626",
+          "627",
+          "628",
+          "629",
+          "630",
+          "631",
+          "632",
+          "633",
+          "634",
+          "635",
+          "636",
+          "637",
+          "638",
+          "639",
+          "640",
+          "641",
+          "642",
+          "643",
+          "644",
+          "645",
+          "646",
+          "647",
+          "648",
+          "649",
+          "650",
+          "651",
+          "652",
+          "653",
+          "654",
+          "655",
+          "656",
+          "657",
+          "658",
+          "659",
+          "660",
+          "661",
+          "662",
+          "663",
+          "664",
+          "665",
+          "666",
+          "667",
+          "668",
+          "669",
+          "670",
+          "671",
+          "672",
+          "673",
+          "674",
+          "675",
+          "676",
+          "677",
+          "678",
+          "679",
+          "680",
+          "681",
+          "682",
+          "683",
+          "684",
+          "685",
+          "686",
+          "687",
+          "688",
+          "689",
+          "690",
+          "691",
+          "692",
+          "693",
+          "694",
+          "695",
+          "696",
+          "697",
+          "698",
+          "699",
+          "700",
+          "701",
+          "702",
+          "703",
+          "704",
+          "705",
+          "706",
+          "707",
+          "708",
+          "709",
+          "710",
+          "711",
+          "712",
+          "713",
+          "714",
+          "715",
+          "716",
+          "717",
+          "718",
+          "719",
+          "720",
+          "721",
+          "722",
+          "723",
+          "724",
+          "725",
+          "726",
+          "727",
+          "728",
+          "729",
+          "730",
+          "731",
+          "732",
+          "733",
+          "734",
+          "735",
+          "736",
+          "737",
+          "738",
+          "739",
+          "740",
+          "741",
+          "742",
+          "743",
+          "744",
+          "745",
+          "746",
+          "747",
+          "748",
+          "749",
+          "750",
+          "751",
+          "752",
+          "753",
+          "754",
+          "755",
+          "756",
+          "757",
+          "758",
+          "759",
+          "760",
+          "761",
+          "762",
+          "763",
+          "764",
+          "765",
+          "766",
+          "767",
+          "768",
+          "769",
+          "770",
+          "771",
+          "772",
+          "773",
+          "774",
+          "775",
+          "776",
+          "777",
+          "778",
+          "779",
+          "780",
+          "781",
+          "782",
+          "783",
+          "784",
+          "785",
+          "786",
+          "787",
+          "788",
+          "789",
+          "790",
+          "791",
+          "792",
+          "793",
+          "794",
+          "795",
+          "796",
+          "797",
+          "798",
+          "799",
+          "800",
+          "801",
+          "802",
+          "803",
+          "804",
+          "805",
+          "806",
+          "807",
+          "808",
+          "809",
+          "810",
+          "811",
+          "812",
+          "813",
+          "814",
+          "815",
+          "816",
+          "817",
+          "818",
+          "819",
+          "820",
+          "821",
+          "822",
+          "823",
+          "824",
+          "825",
+          "826",
+          "827",
+          "828",
+          "829",
+          "830",
+          "831",
+          "832",
+          "833",
+          "834",
+          "835",
+          "836",
+          "837",
+          "838",
+          "839",
+          "840",
+          "841",
+          "842",
+          "843",
+          "844",
+          "845",
+          "846",
+          "847",
+          "848",
+          "849",
+          "850",
+          "851",
+          "852",
+          "853",
+          "854",
+          "855",
+          "856",
+          "857",
+          "858",
+          "859",
+          "860",
+          "861",
+          "862",
+          "863",
+          "864",
+          "865",
+          "866",
+          "867",
+          "868",
+          "869",
+          "870",
+          "871",
+          "872",
+          "873",
+          "874",
+          "875",
+          "876",
+          "877",
+          "878",
+          "879",
+          "880",
+          "881",
+          "882",
+          "883",
+          "884",
+          "885",
+          "886",
+          "887",
+          "888",
+          "889",
+          "890",
+          "891",
+          "892",
+          "893",
+          "894",
+          "895",
+          "896",
+          "897",
+          "898",
+          "899",
+          "900",
+          "901",
+          "902",
+          "903",
+          "904",
+          "905",
+          "906",
+          "907",
+          "908",
+          "909",
+          "910",
+          "911",
+          "912",
+          "913",
+          "914",
+          "915",
+          "916",
+          "917",
+          "918",
+          "919",
+          "920",
+          "921",
+          "922",
+          "923",
+          "924",
+          "925",
+          "926",
+          "927",
+          "928",
+          "929",
+          "930",
+          "931",
+          "932",
+          "933",
+          "934",
+          "935",
+          "936",
+          "937",
+          "938",
+          "939",
+          "940",
+          "941",
+          "942",
+          "943",
+          "944",
+          "945",
+          "946",
+          "947",
+          "948",
+          "949",
+          "950",
+          "951",
+          "952",
+          "953",
+          "954",
+          "955",
+          "956",
+          "957",
+          "958",
+          "959",
+          "960",
+          "961",
+          "962",
+          "963",
+          "964",
+          "965",
+          "966",
+          "967",
+          "968",
+          "969",
+          "970",
+          "971",
+          "972",
+          "973",
+          "974",
+          "975",
+          "976",
+          "977",
+          "978",
+          "979",
+          "980",
+          "981",
+          "982",
+          "983",
+          "984",
+          "985",
+          "986",
+          "987",
+          "988",
+          "989",
+          "990",
+          "991",
+          "992",
+          "993",
+          "994",
+          "995",
+          "996",
+          "997",
+          "998",
+          "999",
+          "1000",
+          "1001",
+          "1002",
+          "1003",
+          "1004",
+          "1005",
+          "1006",
+          "1007",
+          "1008",
+          "1009",
+          "1010",
+          "1011",
+          "1012",
+          "1013",
+          "1014",
+          "1015",
+          "1016",
+          "1017",
+          "1018",
+          "1019",
+          "1020",
+          "1021",
+          "1022",
+          "1023",
+          "1024",
+          "1025",
+          "1026",
+          "1027",
+          "1028",
+          "1029",
+          "1030",
+          "1031",
+          "1032",
+          "1033",
+          "1034",
+          "1035",
+          "1036",
+          "1037",
+          "1038",
+          "1039",
+          "1040",
+          "1041",
+          "1042",
+          "1043",
+          "1044",
+          "1045",
+          "1046",
+          "1047",
+          "1048",
+          "1049",
+          "1050",
+          "1051",
+          "1052",
+          "1053",
+          "1054",
+          "1055",
+          "1056",
+          "1057",
+          "1058",
+          "1059",
+          "1060",
+          "1061",
+          "1062",
+          "1063",
+          "1064",
+          "1065",
+          "1066",
+          "1067",
+          "1068",
+          "1069",
+          "1070",
+          "1071",
+          "1072",
+          "1073",
+          "1074",
+          "1075",
+          "1076",
+          "1077",
+          "1078",
+          "1079",
+          "1080",
+          "1081",
+          "1082",
+          "1083",
+          "1084",
+          "1085",
+          "1086",
+          "1087",
+          "1088",
+          "1089",
+          "1090",
+          "1091",
+          "1092",
+          "1093",
+          "1094",
+          "1095",
+          "1096",
+          "1097",
+          "1098",
+          "1099",
+          "1100",
+          "1101",
+          "1102",
+          "1103",
+          "1104",
+          "1105",
+          "1106",
+          "1107",
+          "1108",
+          "1109",
+          "1110",
+          "1111",
+          "1112",
+          "1113",
+          "1114",
+          "1115",
+          "1116",
+          "1117",
+          "1118",
+          "1119",
+          "1120",
+          "1121",
+          "1122",
+          "1123",
+          "1124",
+          "1125",
+          "1126",
+          "1127",
+          "1128",
+          "1129",
+          "1130",
+          "1131",
+          "1132",
+          "1133",
+          "1134",
+          "1135",
+          "1136",
+          "1137",
+          "1138",
+          "1139",
+          "1140",
+          "1141",
+          "1142",
+          "1143",
+          "1144",
+          "1145",
+          "1146",
+          "1147",
+          "1148",
+          "1149",
+          "1150",
+          "1151",
+          "1152",
+          "1153",
+          "1154",
+          "1155",
+          "1156",
+          "1157",
+          "1158",
+          "1159",
+          "1160",
+          "1161",
+          "1162",
+          "1163",
+          "1164",
+          "1165",
+          "1166",
+          "1167",
+          "1168",
+          "1169",
+          "1170",
+          "1171",
+          "1172",
+          "1173",
+          "1174",
+          "1175",
+          "1176",
+          "1177",
+          "1178",
+          "1179",
+          "1180",
+          "1181",
+          "1182",
+          "1183",
+          "1184",
+          "1185",
+          "1186",
+          "1187",
+          "1188",
+          "1189",
+          "1190",
+          "1191",
+          "1192",
+          "1193",
+          "1194",
+          "1195",
+          "1196",
+          "1197",
+          "1198",
+          "1199",
+          "1200",
+          "1201",
+          "1202",
+          "1203",
+          "1204",
+          "1205",
+          "1206",
+          "1207",
+          "1208",
+          "1209",
+          "1210",
+          "1211",
+          "1212",
+          "1213",
+          "1214",
+          "1215",
+          "1216",
+          "1217",
+          "1218",
+          "1219",
+          "1220",
+          "1221",
+          "1222",
+          "1223",
+          "1224",
+          "1225",
+          "1226",
+          "1227",
+          "1228",
+          "1229",
+          "1230",
+          "1231",
+          "1232",
+          "1233",
+          "1234",
+          "1235",
+          "1236",
+          "1237",
+          "1238",
+          "1239",
+          "1240",
+          "1241",
+          "1242",
+          "1243",
+          "1244",
+          "1245",
+          "1246",
+          "1247",
+          "1248",
+          "1249",
+          "1250",
+          "1251",
+          "1252",
+          "1253",
+          "1254",
+          "1255",
+          "1256",
+          "1257",
+          "1258",
+          "1259",
+          "1260",
+          "1261",
+          "1262",
+          "1263",
+          "1264",
+          "1265",
+          "1266",
+          "1267",
+          "1268",
+          "1269",
+          "1270",
+          "1271",
+          "1272",
+          "1273",
+          "1274",
+          "1275",
+          "1276",
+          "1277",
+          "1278",
+          "1279",
+          "1280",
+          "1281",
+          "1282",
+          "1283",
+          "1284",
+          "1285",
+          "1286",
+          "1287",
+          "1288",
+          "1289",
+          "1290",
+          "1291",
+          "1292",
+          "1293",
+          "1294",
+          "1295",
+          "1296",
+          "1297",
+          "1298",
+          "1299",
+          "1300",
+          "1301",
+          "1302",
+          "1303",
+          "1304",
+          "1305",
+          "1306",
+          "1307",
+          "1308",
+          "1309",
+          "1310",
+          "1311",
+          "1312",
+          "1313",
+          "1314",
+          "1315",
+          "1316",
+          "1317",
+          "1318",
+          "1319",
+          "1320",
+          "1321",
+          "1322",
+          "1323",
+          "1324",
+          "1325",
+          "1326",
+          "1327",
+          "1328",
+          "1329",
+          "1330",
+          "1331",
+          "1332",
+          "1333",
+          "1334",
+          "1335",
+          "1336",
+          "1337",
+          "1338",
+          "1339",
+          "1340",
+          "1341",
+          "1342",
+          "1343",
+          "1344",
+          "1345",
+          "1346",
+          "1347",
+          "1348",
+          "1349",
+          "1350",
+          "1351",
+          "1352",
+          "1353",
+          "1354",
+          "1355",
+          "1356",
+          "1357",
+          "1358",
+          "1359",
+          "1360",
+          "1361",
+          "1362",
+          "1363",
+          "1364",
+          "1365",
+          "1366",
+          "1367",
+          "1368",
+          "1369",
+          "1370",
+          "1371",
+          "1372",
+          "1373",
+          "1374",
+          "1375",
+          "1376",
+          "1377",
+          "1378",
+          "1379",
+          "1380",
+          "1381",
+          "1382",
+          "1383",
+          "1384",
+          "1385",
+          "1386",
+          "1387",
+          "1388",
+          "1389",
+          "1390",
+          "1391",
+          "1392",
+          "1393",
+          "1394",
+          "1395",
+          "1396",
+          "1397",
+          "1398",
+          "1399",
+          "1400",
+          "1401",
+          "1402",
+          "1403",
+          "1404",
+          "1405",
+          "1406",
+          "1407",
+          "1408",
+          "1409",
+          "1410",
+          "1411",
+          "1412",
+          "1413",
+          "1414",
+          "1415",
+          "1416",
+          "1417",
+          "1418",
+          "1419",
+          "1420",
+          "1421",
+          "1422",
+          "1423",
+          "1424",
+          "1425",
+          "1426",
+          "1427",
+          "1428",
+          "1429",
+          "1430",
+          "1431",
+          "1432",
+          "1433",
+          "1434",
+          "1435",
+          "1436",
+          "1437",
+          "1438",
+          "1439",
+          "1440",
+          "1441",
+          "1442",
+          "1443",
+          "1444",
+          "1445",
+          "1446",
+          "1447",
+          "1448",
+          "1449",
+          "1450",
+          "1451",
+          "1452",
+          "1453",
+          "1454",
+          "1455",
+          "1456",
+          "1457",
+          "1458",
+          "1459",
+          "1460",
+          "1461",
+          "1462",
+          "1463",
+          "1464",
+          "1465",
+          "1466",
+          "1467",
+          "1468",
+          "1469",
+          "1470",
+          "1471",
+          "1472",
+          "1473",
+          "1474",
+          "1475",
+          "1476",
+          "1477",
+          "1478",
+          "1479",
+          "1480",
+          "1481",
+          "1482",
+          "1483",
+          "1484",
+          "1485",
+          "1486",
+          "1487",
+          "1488",
+          "1489",
+          "1490",
+          "1491",
+          "1492",
+          "1493",
+          "1494",
+          "1495",
+          "1496",
+          "1497",
+          "1498",
+          "1499",
+          "1500",
+          "1501",
+          "1502",
+          "1503",
+          "1504",
+          "1505",
+          "1506",
+          "1507",
+          "1508",
+          "1509",
+          "1510",
+          "1511",
+          "1512",
+          "1513",
+          "1514",
+          "1515",
+          "1516",
+          "1517",
+          "1518",
+          "1519",
+          "1520",
+          "1521",
+          "1522",
+          "1523",
+          "1524",
+          "1525",
+          "1526",
+          "1527",
+          "1528",
+          "1529",
+          "1530",
+          "1531",
+          "1532",
+          "1533",
+          "1534",
+          "1535",
+          "1536",
+          "1537",
+          "1538",
+          "1539",
+          "1540",
+          "1541",
+          "1542",
+          "1543",
+          "1544",
+          "1545",
+          "1546",
+          "1547",
+          "1548",
+          "1549",
+          "1550",
+          "1551",
+          "1552",
+          "1553",
+          "1554",
+          "1555",
+          "1556",
+          "1557",
+          "1558",
+          "1559",
+          "1560",
+          "1561",
+          "1562",
+          "1563",
+          "1564",
+          "1565",
+          "1566",
+          "1567",
+          "1568",
+          "1569",
+          "1570",
+          "1571",
+          "1572",
+          "1573",
+          "1574",
+          "1575",
+          "1576",
+          "1577",
+          "1578",
+          "1579",
+          "1580",
+          "1581",
+          "1582",
+          "1583",
+          "1584",
+          "1585",
+          "1586",
+          "1587",
+          "1588",
+          "1589",
+          "1590",
+          "1591",
+          "1592",
+          "1593",
+          "1594",
+          "1595",
+          "1596",
+          "1597",
+          "1598",
+          "1599",
+          "1600",
+          "1601",
+          "1602",
+          "1603",
+          "1604",
+          "1605",
+          "1606",
+          "1607",
+          "1608",
+          "1609",
+          "1610",
+          "1611",
+          "1612",
+          "1613",
+          "1614",
+          "1615",
+          "1616",
+          "1617",
+          "1618",
+          "1619",
+          "1620",
+          "1621",
+          "1622",
+          "1623",
+          "1624",
+          "1625",
+          "1626",
+          "1627",
+          "1628",
+          "1629",
+          "1630",
+          "1631",
+          "1632",
+          "1633",
+          "1634",
+          "1635",
+          "1636",
+          "1637",
+          "1638",
+          "1639",
+          "1640",
+          "1641",
+          "1642",
+          "1643",
+          "1644",
+          "1645",
+          "1646",
+          "1647",
+          "1648",
+          "1649",
+          "1650",
+          "1651",
+          "1652",
+          "1653",
+          "1654",
+          "1655",
+          "1656",
+          "1657",
+          "1658",
+          "1659",
+          "1660",
+          "1661",
+          "1662",
+          "1663",
+          "1664",
+          "1665",
+          "1666",
+          "1667",
+          "1668",
+          "1669",
+          "1670",
+          "1671",
+          "1672",
+          "1673",
+          "1674",
+          "1675",
+          "1676",
+          "1677",
+          "1678",
+          "1679",
+          "1680",
+          "1681",
+          "1682",
+          "1683",
+          "1684",
+          "1685",
+          "1686",
+          "1687",
+          "1688",
+          "1689",
+          "1690",
+          "1691",
+          "1692",
+          "1693",
+          "1694",
+          "1695",
+          "1696",
+          "1697",
+          "1698",
+          "1699",
+          "1700",
+          "1701",
+          "1702",
+          "1703",
+          "1704",
+          "1705",
+          "1706",
+          "1707",
+          "1708",
+          "1709",
+          "1710",
+          "1711",
+          "1712",
+          "1713",
+          "1714",
+          "1715",
+          "1716",
+          "1717",
+          "1718",
+          "1719",
+          "1720",
+          "1721",
+          "1722",
+          "1723",
+          "1724",
+          "1725",
+          "1726",
+          "1727",
+          "1728",
+          "1729",
+          "1730",
+          "1731",
+          "1732",
+          "1733",
+          "1734",
+          "1735",
+          "1736",
+          "1737",
+          "1738",
+          "1739",
+          "1740",
+          "1741",
+          "1742",
+          "1743",
+          "1744",
+          "1745",
+          "1746",
+          "1747",
+          "1748",
+          "1749",
+          "1750",
+          "1751",
+          "1752",
+          "1753",
+          "1754",
+          "1755",
+          "1756",
+          "1757",
+          "1758",
+          "1759",
+          "1760",
+          "1761",
+          "1762",
+          "1763",
+          "1764",
+          "1765",
+          "1766",
+          "1767",
+          "1768",
+          "1769",
+          "1770",
+          "1771",
+          "1772",
+          "1773",
+          "1774",
+          "1775",
+          "1776",
+          "1777",
+          "1778",
+          "1779",
+          "1780",
+          "1781",
+          "1782",
+          "1783",
+          "1784",
+          "1785",
+          "1786",
+          "1787",
+          "1788",
+          "1789",
+          "1790",
+          "1791",
+          "1792",
+          "1793",
+          "1794",
+          "1795",
+          "1796",
+          "1797",
+          "1798",
+          "1799",
+          "1800",
+          "1801",
+          "1802",
+          "1803",
+          "1804",
+          "1805",
+          "1806",
+          "1807",
+          "1808",
+          "1809",
+          "1810",
+          "1811",
+          "1812",
+          "1813",
+          "1814",
+          "1815",
+          "1816",
+          "1817",
+          "1818",
+          "1819",
+          "1820",
+          "1821",
+          "1822",
+          "1823",
+          "1824",
+          "1825",
+          "1826",
+          "1827",
+          "1828",
+          "1829",
+          "1830",
+          "1831",
+          "1832",
+          "1833",
+          "1834",
+          "1835",
+          "1836",
+          "1837",
+          "1838"
+        ],
+        "additional_details": {
+          "shuffled": "False"
+        }
+      },
+      "evaluation_timestamp": "1742646860.0",
+      "metric_config": {
+        "evaluation_description": "accuracy",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "additional_details": {
+          "raw_evaluation_name": "accuracy on inspect_evals/piqa for scorer choice"
+        },
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.9238302502720348,
+        "uncertainty": {
+          "standard_error": {
+            "value": 0.0061891751848599675
+          },
+          "num_samples": 1838
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "reasoning": false,
+          "agentic_eval_config": {
+            "available_tools": []
+          },
+          "eval_plan": {
+            "name": "plan",
+            "steps": [
+              "{\"solver\": \"multiple_choice\", \"params\": {\"template\": \"The entire content of your response should be of the following format: 'ANSWER:\\n$LETTER' (without quotes) where LETTER is one of {letters}.\\n\\nGiven either a question or a statement followed by two possible solutions\\nlabelled A and B, choose the most appropriate solution. If a question is given,\\nthe solutions answer the question. If a statement is given, the solutions\\nexplain how to achieve the statement.\\n\\n{question}\\n\\n{choices}\"}, \"params_passed\": {\"template\": \"The entire content of your response should be of the following format: 'ANSWER:\\n$LETTER' (without quotes) where LETTER is one of {letters}.\\n\\nGiven either a question or a statement followed by two possible solutions\\nlabelled A and B, choose the most appropriate solution. If a question is given,\\nthe solutions answer the question. If a statement is given, the solutions\\nexplain how to achieve the statement.\\n\\n{question}\\n\\n{choices}\"}}"
+            ],
+            "config": {}
+          },
+          "eval_limits": {},
+          "sandbox": {}
+        }
+      },
+      "evaluation_result_id": "piqa/mistral_mistral-large-latest/1742646860.0#piqa#accuracy"
+    }
+  ],
+  "detailed_evaluation_results": {
+    "format": "jsonl",
+    "file_path": "./0039055f-58eb-4aa9-a1c7-2832224f611a_samples.jsonl",
+    "hash_algorithm": "sha256",
+    "checksum": "d58cf4e7da5aeceb2f54d7ee28f1e289d9b98bccf1c4a5cbb0d1d1a837a7dc32",
+    "total_rows": 1838
+  }
+}

flat/objects/00/39/00397e7e-edc5-4c97-ae8a-b38b8cf13612.json

@@ -0,0 +1,208 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "ChartX/CogAgent/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Shanghai Artificial Intelligence Laboratory",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "CogAgent",
+    "name": "CogAgent",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "ChartX",
+      "source_data": {
+        "dataset_name": "ChartX",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12185"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the ability of models to extract underlying numerical and textual data from chart images on the ChartX benchmark. The AP@High metric uses the Structuring Chart-oriented Representation Metric (SCRM) with a high tolerance for errors, making it a more lenient evaluation of this perception task. This task is considered pivotal for the interpretability of subsequent reasoning. Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Structural Extraction (AP@High)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Structural Extraction Performance on ChartX (High Tolerance)"
+        },
+        "metric_id": "structural_extraction_performance_on_chartx_high_tolerance",
+        "metric_name": "Structural Extraction Performance on ChartX (High Tolerance)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 6.36
+      },
+      "evaluation_result_id": "ChartX/CogAgent/1771591481.616601#chartx#structural_extraction_performance_on_chartx_high_tolerance"
+    },
+    {
+      "evaluation_name": "ChartX",
+      "source_data": {
+        "dataset_name": "ChartX",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12185"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the ability of models to generate functional Python code that can recreate a given chart image. GPT-4 rates the generated code on a scale from 0 to 5 (GPT-score) based on functionality and visual fidelity. Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Chart Redrawing (GPT-score)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Chart Redrawing Code Generation Quality on ChartX"
+        },
+        "metric_id": "chart_redrawing_code_generation_quality_on_chartx",
+        "metric_name": "Chart Redrawing Code Generation Quality on ChartX",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.48
+      },
+      "evaluation_result_id": "ChartX/CogAgent/1771591481.616601#chartx#chart_redrawing_code_generation_quality_on_chartx"
+    },
+    {
+      "evaluation_name": "ChartX",
+      "source_data": {
+        "dataset_name": "ChartX",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12185"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the quality of concise summaries generated by models, which should highlight key trends and insights from the chart data on the ChartX benchmark. GPT-4 rates the outputs on a scale from 0 to 5 (GPT-score). Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Chart Summarization (GPT-score)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Chart Summarization Quality on ChartX"
+        },
+        "metric_id": "chart_summarization_quality_on_chartx",
+        "metric_name": "Chart Summarization Quality on ChartX",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.61
+      },
+      "evaluation_result_id": "ChartX/CogAgent/1771591481.616601#chartx#chart_summarization_quality_on_chartx"
+    },
+    {
+      "evaluation_name": "ChartX",
+      "source_data": {
+        "dataset_name": "ChartX",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12185"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the ability of models to accurately determine the specific type of a given chart (e.g., \"bar chart\") on the ChartX benchmark. Performance is measured by Exact Match (EM) accuracy. Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Chart Type Classification (EM %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Chart Type Classification on ChartX"
+        },
+        "metric_id": "chart_type_classification_on_chartx",
+        "metric_name": "Chart Type Classification on ChartX",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 61.11
+      },
+      "evaluation_result_id": "ChartX/CogAgent/1771591481.616601#chartx#chart_type_classification_on_chartx"
+    },
+    {
+      "evaluation_name": "ChartX",
+      "source_data": {
+        "dataset_name": "ChartX",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12185"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the ability of models to answer questions that can be solely derived from the chart's visual and underlying data on the ChartX benchmark. Performance is evaluated using GPT-4 (GPT-acc), which assesses the correctness of answers against ground truth, allowing for a 5% margin of error for numerical responses. Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Question Answering (GPT-acc %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Question Answering Performance on ChartX"
+        },
+        "metric_id": "question_answering_performance_on_chartx",
+        "metric_name": "Question Answering Performance on ChartX",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 25.95
+      },
+      "evaluation_result_id": "ChartX/CogAgent/1771591481.616601#chartx#question_answering_performance_on_chartx"
+    },
+    {
+      "evaluation_name": "ChartX",
+      "source_data": {
+        "dataset_name": "ChartX",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2402.12185"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the ability of models to extract underlying data from chart images on the ChartX benchmark. The AP@Strict metric uses the Structuring Chart-oriented Representation Metric (SCRM) with a strict (zero) tolerance for errors, making it the most challenging evaluation of this perception task. Higher scores are better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Structural Extraction (AP@Strict)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Structural Extraction Performance on ChartX (Strict Tolerance)"
+        },
+        "metric_id": "structural_extraction_performance_on_chartx_strict_tolerance",
+        "metric_name": "Structural Extraction Performance on ChartX (Strict Tolerance)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 2.89
+      },
+      "evaluation_result_id": "ChartX/CogAgent/1771591481.616601#chartx#structural_extraction_performance_on_chartx_strict_tolerance"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/3a/003a5528-7942-4b1e-a181-42f71ad8fff7.json

@@ -0,0 +1,388 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "MATP-BENCH/Llama3.2-V/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "HKUST",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Llama3.2-V",
+    "name": "Llama3.2-V",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the ability of Multimodal Large Language Models (MLLMs) to perform end-to-end multimodal automated theorem proving (Task 1) in the Coq formal language. The metric is pass@10, representing the percentage of problems for which at least one of 10 generated proof candidates was formally verifiable. This result is for the overall MATP-BENCH dataset.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@10 (%) - Task 1 (Coq)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Coq (pass@10)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_coq_pass_10",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Coq (pass@10)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 7.37
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_coq_pass_10"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates end-to-end multimodal automated theorem proving (Task 1) in Coq. The metric is pass@5, representing the percentage of problems for which at least one of 5 generated proof candidates was formally verifiable. These supplementary results show performance with fewer generation attempts compared to pass@10.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@5 (%) - Task 1 (Coq)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Coq (pass@5)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_coq_pass_5",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Coq (pass@5)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 3.4
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_coq_pass_5"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates end-to-end multimodal automated theorem proving (Task 1) in Isabelle under the strictest condition. The metric is pass@1, representing the percentage of problems solved on the first attempt. These supplementary results highlight the extreme difficulty of the task.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@1 (%) - Task 1 (Isabelle)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Isabelle (pass@1)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_isabelle_pass_1",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Isabelle (pass@1)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.71
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_isabelle_pass_1"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the ability of Multimodal Large Language Models (MLLMs) to perform end-to-end multimodal automated theorem proving (Task 1) in the Isabelle formal language. The metric is pass@10, representing the percentage of problems for which at least one of 10 generated proof candidates was formally verifiable. This result is for the overall MATP-BENCH dataset.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@10 (%) - Task 1 (Isabelle)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Isabelle (pass@10)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_isabelle_pass_10",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Isabelle (pass@10)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 2.45
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_isabelle_pass_10"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates end-to-end multimodal automated theorem proving (Task 1) in Isabelle. The metric is pass@5, representing the percentage of problems for which at least one of 5 generated proof candidates was formally verifiable. These supplementary results show performance with fewer generation attempts compared to pass@10.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@5 (%) - Task 1 (Isabelle)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Isabelle (pass@5)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_isabelle_pass_5",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Isabelle (pass@5)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.42
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_isabelle_pass_5"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates end-to-end multimodal automated theorem proving (Task 1) in Lean 4 under the strictest condition. The metric is pass@1, representing the percentage of problems solved on the first attempt. These supplementary results highlight the extreme difficulty of the task.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@1 (%) - Task 1 (Lean 4)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Lean 4 (pass@1)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_lean_4_pass_1",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Lean 4 (pass@1)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.95
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_lean_4_pass_1"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the ability of Multimodal Large Language Models (MLLMs) to perform end-to-end multimodal automated theorem proving (Task 1) in the Lean 4 formal language. The metric is pass@10, representing the percentage of problems for which at least one of 10 generated proof candidates was formally verifiable. This result is for the overall MATP-BENCH dataset.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@10 (%) - Task 1 (Lean 4)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Lean 4 (pass@10)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_lean_4_pass_10",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Lean 4 (pass@10)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 2.46
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_lean_4_pass_10"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates end-to-end multimodal automated theorem proving (Task 1) in Lean 4. The metric is pass@5, representing the percentage of problems for which at least one of 5 generated proof candidates was formally verifiable. These supplementary results show performance with fewer generation attempts compared to pass@10.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@5 (%) - Task 1 (Lean 4)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Lean 4 (pass@5)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_lean_4_pass_5",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Lean 4 (pass@5)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 1.61
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_lean_4_pass_5"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the ability of MLLMs to correctly formalize a theorem from multimodal input (Task 2) into the Coq formal language. Success is judged by GPT-4o. The metric is pass@10, representing success if at least one of 10 formalizations is correct. The high scores compared to Task 1 highlight that proof generation, not formalization, is the main bottleneck.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@10 (%) - Task 2 (Coq)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Multimodal Theorem Formalization in Coq (pass@10)"
+        },
+        "metric_id": "multimodal_theorem_formalization_in_coq_pass_10",
+        "metric_name": "Multimodal Theorem Formalization in Coq (pass@10)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 15.97
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#multimodal_theorem_formalization_in_coq_pass_10"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the ability of MLLMs to correctly formalize a theorem from multimodal input (Task 2) into the Isabelle formal language. Success is judged by GPT-4o. The metric is pass@10, representing success if at least one of 10 formalizations is correct. The high scores compared to Task 1 highlight that proof generation, not formalization, is the main bottleneck.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@10 (%) - Task 2 (Isabelle)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Multimodal Theorem Formalization in Isabelle (pass@10)"
+        },
+        "metric_id": "multimodal_theorem_formalization_in_isabelle_pass_10",
+        "metric_name": "Multimodal Theorem Formalization in Isabelle (pass@10)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 20.52
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#multimodal_theorem_formalization_in_isabelle_pass_10"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates end-to-end multimodal automated theorem proving (Task 1) in Coq under the strictest condition. The metric is pass@1, representing the percentage of problems solved on the first attempt. These supplementary results highlight the extreme difficulty of the task.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@1 (%) - Task 1 (Coq)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "End-to-End Multimodal Theorem Proving in Coq (pass@1)"
+        },
+        "metric_id": "end_to_end_multimodal_theorem_proving_in_coq_pass_1",
+        "metric_name": "End-to-End Multimodal Theorem Proving in Coq (pass@1)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 2.08
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#end_to_end_multimodal_theorem_proving_in_coq_pass_1"
+    },
+    {
+      "evaluation_name": "MATP-BENCH",
+      "source_data": {
+        "dataset_name": "MATP-BENCH",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2506.06034"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the ability of MLLMs to correctly formalize a theorem from multimodal input (Task 2) into the Lean 4 formal language. Success is judged by GPT-4o. The metric is pass@10, representing success if at least one of 10 formalizations is correct. The high scores compared to Task 1 highlight that proof generation, not formalization, is the main bottleneck.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@10 (%) - Task 2 (Lean 4)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Multimodal Theorem Formalization in Lean 4 (pass@10)"
+        },
+        "metric_id": "multimodal_theorem_formalization_in_lean_4_pass_10",
+        "metric_name": "Multimodal Theorem Formalization in Lean 4 (pass@10)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 19.72
+      },
+      "evaluation_result_id": "MATP-BENCH/Llama3.2-V/1771591481.616601#matp_bench#multimodal_theorem_formalization_in_lean_4_pass_10"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/3b/003bd91c-f5eb-4baa-a677-8579ca981e7a.json

@@ -0,0 +1,148 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "YouCookII-TVS/VideoTree/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "University of Illinois at Urbana-Champaign",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "VideoTree",
+    "name": "VideoTree",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "YouCookII-TVS",
+      "source_data": {
+        "dataset_name": "YouCookII-TVS",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.21094"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures VideoQA accuracy on the YouCookII-TVS benchmark, which focuses on temporal reasoning in instructional cooking videos. The evaluation is performed with the paper's proposed Temporal Visual Screening (TVS) process applied as a pre-processing step during inference. This benchmark is newly introduced in the paper.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "VideoQA Accuracy on YouCookII-TVS Benchmark with TVS"
+        },
+        "metric_id": "videoqa_accuracy_on_youcookii_tvs_benchmark_with_tvs",
+        "metric_name": "VideoQA Accuracy on YouCookII-TVS Benchmark with TVS",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 69.4
+      },
+      "evaluation_result_id": "YouCookII-TVS/VideoTree/1771591481.616601#youcookii_tvs#videoqa_accuracy_on_youcookii_tvs_benchmark_with_tvs"
+    },
+    {
+      "evaluation_name": "YouCookII-TVS",
+      "source_data": {
+        "dataset_name": "YouCookII-TVS",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.21094"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures VideoQA accuracy on the ActivityNetQA benchmark, which contains question-answer pairs about activities in user-generated videos from YouTube. The evaluation is performed with the paper's proposed Temporal Visual Screening (TVS) process applied as a pre-processing step during inference.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "VideoQA Accuracy on ActivityNetQA Benchmark with TVS"
+        },
+        "metric_id": "videoqa_accuracy_on_activitynetqa_benchmark_with_tvs",
+        "metric_name": "VideoQA Accuracy on ActivityNetQA Benchmark with TVS",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 63.6
+      },
+      "evaluation_result_id": "YouCookII-TVS/VideoTree/1771591481.616601#youcookii_tvs#videoqa_accuracy_on_activitynetqa_benchmark_with_tvs"
+    },
+    {
+      "evaluation_name": "YouCookII-TVS",
+      "source_data": {
+        "dataset_name": "YouCookII-TVS",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.21094"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures open-ended VideoQA accuracy on the NExT-OE benchmark, designed to evaluate complex event understanding, including temporal, causal, and descriptive reasoning. The evaluation is performed with the paper's proposed Temporal Visual Screening (TVS) process applied as a pre-processing step during inference.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "VideoQA Accuracy on NExT-OE Benchmark with TVS"
+        },
+        "metric_id": "videoqa_accuracy_on_next_oe_benchmark_with_tvs",
+        "metric_name": "VideoQA Accuracy on NExT-OE Benchmark with TVS",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 61.9
+      },
+      "evaluation_result_id": "YouCookII-TVS/VideoTree/1771591481.616601#youcookii_tvs#videoqa_accuracy_on_next_oe_benchmark_with_tvs"
+    },
+    {
+      "evaluation_name": "YouCookII-TVS",
+      "source_data": {
+        "dataset_name": "YouCookII-TVS",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.21094"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures multiple-choice VideoQA accuracy on the NExT-QA benchmark, designed to evaluate complex event understanding including temporal, causal, and descriptive reasoning. The evaluation is performed with the paper's proposed Temporal Visual Screening (TVS) process applied as a pre-processing step during inference. Note that proprietary models like GPT-4 series were not evaluated on this specific benchmark in the paper.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "VideoQA Accuracy on NExT-QA Benchmark with TVS"
+        },
+        "metric_id": "videoqa_accuracy_on_next_qa_benchmark_with_tvs",
+        "metric_name": "VideoQA Accuracy on NExT-QA Benchmark with TVS",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 65.2
+      },
+      "evaluation_result_id": "YouCookII-TVS/VideoTree/1771591481.616601#youcookii_tvs#videoqa_accuracy_on_next_qa_benchmark_with_tvs"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/3c/003c01cc-1a1a-46a2-a2eb-7462ec972c9b.json

@@ -0,0 +1,118 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "CFVBench/claude-opus-4/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Chinese Academy of Sciences",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "claude-opus-4",
+    "name": "claude-opus-4",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "CFVBench",
+      "source_data": {
+        "dataset_name": "CFVBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09266"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the zero-shot generation performance of Multimodal Large Language Models (MLLMs) on CFVBench using the F1-score, which is the harmonic mean of keypoint-based precision and recall. This score provides a balanced assessment of a model's ability to generate factually accurate and complete answers.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1-score",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Baseline Generation Performance on CFVBench (F1-score)"
+        },
+        "metric_id": "baseline_generation_performance_on_cfvbench_f1_score",
+        "metric_name": "Baseline Generation Performance on CFVBench (F1-score)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.185
+      },
+      "evaluation_result_id": "CFVBench/claude-opus-4/1771591481.616601#cfvbench#baseline_generation_performance_on_cfvbench_f1_score"
+    },
+    {
+      "evaluation_name": "CFVBench",
+      "source_data": {
+        "dataset_name": "CFVBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09266"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the overall quality of generated answers from MLLMs on a 1-5 Likert scale, as judged by other large language models (Qwen3-8B-Instruct and GLM-4-9B). This score provides a holistic assessment of answer quality.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Likert Score (1-5)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Baseline Generation Performance on CFVBench (LLM-as-Judge Likert Score)"
+        },
+        "metric_id": "baseline_generation_performance_on_cfvbench_llm_as_judge_likert_score",
+        "metric_name": "Baseline Generation Performance on CFVBench (LLM-as-Judge Likert Score)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 3.2
+      },
+      "evaluation_result_id": "CFVBench/claude-opus-4/1771591481.616601#cfvbench#baseline_generation_performance_on_cfvbench_llm_as_judge_likert_score"
+    },
+    {
+      "evaluation_name": "CFVBench",
+      "source_data": {
+        "dataset_name": "CFVBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09266"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Evaluates the zero-shot generation performance of MLLMs on CFVBench using overall keypoint-based recall. This metric measures the proportion of correctly identified textual and visual keypoints from the ground truth that are present in the generated answer.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Recall",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Baseline Generation Performance on CFVBench (Overall Recall)"
+        },
+        "metric_id": "baseline_generation_performance_on_cfvbench_overall_recall",
+        "metric_name": "Baseline Generation Performance on CFVBench (Overall Recall)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.2079
+      },
+      "evaluation_result_id": "CFVBench/claude-opus-4/1771591481.616601#cfvbench#baseline_generation_performance_on_cfvbench_overall_recall"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/3d/003d8290-b0b3-4bcf-8578-8e65f31e6603.json

@@ -0,0 +1,268 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "New York University",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)",
+    "name": "Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "LiveBench",
+      "source_data": {
+        "dataset_name": "LiveBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2406.19314"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the standard Coding category of LiveBench, involving code generation and completion tasks based on recent Leetcode and AtCoder problems from LiveCodeBench. LiveBench is a contamination-limited benchmark with frequently updated questions and objective ground-truth scoring.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Coding Average (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "LiveBench - Coding"
+        },
+        "metric_id": "livebench_coding",
+        "metric_name": "LiveBench - Coding",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 66.41
+      },
+      "evaluation_result_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601#livebench#livebench_coding"
+    },
+    {
+      "evaluation_name": "LiveBench",
+      "source_data": {
+        "dataset_name": "LiveBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2406.19314"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Data Analysis category of LiveBench, which tests models on tasks like column type annotation, table reformatting between different schemas, and predicting join columns between two tables, using recent Kaggle and Socrata datasets. LiveBench is a contamination-limited benchmark with frequently updated questions and objective ground-truth scoring.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Data Analysis Average (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "LiveBench - Data Analysis"
+        },
+        "metric_id": "livebench_data_analysis",
+        "metric_name": "LiveBench - Data Analysis",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 67.25
+      },
+      "evaluation_result_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601#livebench#livebench_data_analysis"
+    },
+    {
+      "evaluation_name": "LiveBench",
+      "source_data": {
+        "dataset_name": "LiveBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2406.19314"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall performance score on the LiveBench benchmark, calculated as the average of the seven category scores (Reasoning, Coding, Agentic Coding, Mathematics, Data Analysis, Language, and Instruction Following). LiveBench is a contamination-limited benchmark with frequently updated questions and objective ground-truth scoring.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Global Average (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "LiveBench - Global Average"
+        },
+        "metric_id": "livebench_global_average",
+        "metric_name": "LiveBench - Global Average",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 58.42
+      },
+      "evaluation_result_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601#livebench#livebench_global_average"
+    },
+    {
+      "evaluation_name": "LiveBench",
+      "source_data": {
+        "dataset_name": "LiveBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2406.19314"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Instruction Following (IF) category of LiveBench. Models must perform a primary task (e.g., summarize, paraphrase) on a recent news article while adhering to multiple verifiable constraints like word count or keyword inclusion. LiveBench is a contamination-limited benchmark with frequently updated questions and objective ground-truth scoring.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Instruction Following Average (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "LiveBench - Instruction Following (IF)"
+        },
+        "metric_id": "livebench_instruction_following_if",
+        "metric_name": "LiveBench - Instruction Following (IF)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 80.74
+      },
+      "evaluation_result_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601#livebench#livebench_instruction_following_if"
+    },
+    {
+      "evaluation_name": "LiveBench",
+      "source_data": {
+        "dataset_name": "LiveBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2406.19314"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Language Comprehension category of LiveBench. Tasks include sorting words into groups based on common connections, fixing typos in recent ArXiv abstracts while preserving style, and reordering shuffled sentences of movie synopses. LiveBench is a contamination-limited benchmark with frequently updated questions and objective ground-truth scoring.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Language Average (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "LiveBench - Language Comprehension"
+        },
+        "metric_id": "livebench_language_comprehension",
+        "metric_name": "LiveBench - Language Comprehension",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 47.93
+      },
+      "evaluation_result_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601#livebench#livebench_language_comprehension"
+    },
+    {
+      "evaluation_name": "LiveBench",
+      "source_data": {
+        "dataset_name": "LiveBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2406.19314"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Mathematics category of LiveBench, featuring questions from recent math competitions (AMC, AIME), Olympiad-level proof ordering, and harder synthetically generated problems (AMPS_Hard). LiveBench is a contamination-limited benchmark with frequently updated questions and objective ground-truth scoring.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Mathematics Average (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "LiveBench - Mathematics"
+        },
+        "metric_id": "livebench_mathematics",
+        "metric_name": "LiveBench - Mathematics",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 75.9
+      },
+      "evaluation_result_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601#livebench#livebench_mathematics"
+    },
+    {
+      "evaluation_name": "LiveBench",
+      "source_data": {
+        "dataset_name": "LiveBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2406.19314"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Agentic Coding category of LiveBench, a novel task where models must solve real-world repository issues in a real development environment, testing their ability to act as autonomous agents. LiveBench is a contamination-limited benchmark with frequently updated questions and objective ground-truth scoring.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Agentic Coding Average (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "LiveBench - Agentic Coding"
+        },
+        "metric_id": "livebench_agentic_coding",
+        "metric_name": "LiveBench - Agentic Coding",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 5
+      },
+      "evaluation_result_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601#livebench#livebench_agentic_coding"
+    },
+    {
+      "evaluation_name": "LiveBench",
+      "source_data": {
+        "dataset_name": "LiveBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2406.19314"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Reasoning category of LiveBench, which includes tasks like complex Boolean logic puzzles (Web of Lies v2), classic logic deduction (Zebra Puzzles), and spatial reasoning with 2D/3D shapes. LiveBench is a contamination-limited benchmark with frequently updated questions and objective ground-truth scoring.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Reasoning Average (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "LiveBench - Reasoning"
+        },
+        "metric_id": "livebench_reasoning",
+        "metric_name": "LiveBench - Reasoning",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 65.72
+      },
+      "evaluation_result_id": "LiveBench/Gemini 2.5 Flash Lite (Max Thinking) (2025-06-17)/1771591481.616601#livebench#livebench_reasoning"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/3f/003f2eaa-3649-4ea2-82c7-8b8f86c92a0b.json

@@ -0,0 +1,388 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "ReasonBench/InternVL2/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Beijing Electronic Science & Technology Institute",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "InternVL2",
+    "name": "InternVL2",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The overall average accuracy of Visual Language Models (VLMs) across all 29 tasks in the ReasonBench benchmark using an integrated input format. This metric serves as the primary measure of a model's general capability in complex graphic reasoning. Scores are based on Pass@1 single-attempt evaluation.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Overall Accuracy on ReasonBench"
+        },
+        "metric_id": "overall_accuracy_on_reasonbench",
+        "metric_name": "Overall Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 23.62
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#overall_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy on tasks involving attribute and quantitative reasoning, requiring models to detect numerical patterns in lines, surfaces, points, and elements.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Attribute & Quantitative Reasoning Accuracy on ReasonBench"
+        },
+        "metric_id": "attribute_quantitative_reasoning_accuracy_on_reasonbench",
+        "metric_name": "Attribute & Quantitative Reasoning Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 23.51
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#attribute_quantitative_reasoning_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on tasks evaluating the arrangement, exchange, and transformation of black and white blocks. This tests a VLM's ability to infer logical transformations within binary visual structures.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Black & White Block Reasoning Accuracy on ReasonBench"
+        },
+        "metric_id": "black_white_block_reasoning_accuracy_on_reasonbench",
+        "metric_name": "Black & White Block Reasoning Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 18.75
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#black_white_block_reasoning_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy on subsets derived from Mensa Intelligence Tests. This provides a direct benchmark comparison against human intelligence assessment metrics for tasks with increasing difficulty.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Mensa Test Accuracy on ReasonBench"
+        },
+        "metric_id": "mensa_test_accuracy_on_reasonbench",
+        "metric_name": "Mensa Test Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 9.42
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#mensa_test_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on tasks that combine multiple reasoning rules, requiring the model to apply diverse inference strategies in a single problem. This evaluates a VLM's ability to handle complex, multi-rule reasoning scenarios.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Miscellaneous Reasoning Accuracy on ReasonBench"
+        },
+        "metric_id": "miscellaneous_reasoning_accuracy_on_reasonbench",
+        "metric_name": "Miscellaneous Reasoning Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 20.59
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#miscellaneous_reasoning_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The overall average accuracy of a subset of VLMs across all 29 tasks in the ReasonBench benchmark using a separated input format, where the problem and answer choices are provided as separate images. This measures performance under an alternative input modality.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Accuracy on ReasonBench (Separated Format)"
+        },
+        "metric_id": "overall_accuracy_on_reasonbench_separated_format",
+        "metric_name": "Overall Accuracy on ReasonBench (Separated Format)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 24.05
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#overall_accuracy_on_reasonbench_separated_format"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy on tasks evaluating positional reasoning, including translation, rotation, and their combination. This dimension assesses if VLMs can accurately recognize movement patterns and infer spatial transformations.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Positional Reasoning Accuracy on ReasonBench"
+        },
+        "metric_id": "positional_reasoning_accuracy_on_reasonbench",
+        "metric_name": "Positional Reasoning Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 28.42
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#positional_reasoning_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy on subsets derived from Raven’s Progressive Matrices. These tasks are designed to evaluate abstract reasoning capabilities.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Raven's Progressive Matrices Accuracy on ReasonBench"
+        },
+        "metric_id": "raven_s_progressive_matrices_accuracy_on_reasonbench",
+        "metric_name": "Raven's Progressive Matrices Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 20.84
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#raven_s_progressive_matrices_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy on tasks evaluating spatial modeling and 3D reasoning, including cube matching, 3D assembly, three-view drawings, and sectional views. This dimension was the most challenging for the human baseline.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Spatial Reasoning Accuracy on ReasonBench"
+        },
+        "metric_id": "spatial_reasoning_accuracy_on_reasonbench",
+        "metric_name": "Spatial Reasoning Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 26.02
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#spatial_reasoning_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy on special reasoning tasks, including 2D composition and inter-figure relations (e.g., intersections, unions). This dimension assesses a VLM's ability to perform logical set operations on visual elements.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Special Reasoning Accuracy on ReasonBench"
+        },
+        "metric_id": "special_reasoning_accuracy_on_reasonbench",
+        "metric_name": "Special Reasoning Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 30.61
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#special_reasoning_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on tasks that involve recognizing letter and number sequences that follow specific patterns, evaluating a VLM's ability to infer the next element based on explicit rules.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Alphanumeric Reasoning Accuracy on ReasonBench"
+        },
+        "metric_id": "alphanumeric_reasoning_accuracy_on_reasonbench",
+        "metric_name": "Alphanumeric Reasoning Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 22.22
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#alphanumeric_reasoning_accuracy_on_reasonbench"
+    },
+    {
+      "evaluation_name": "ReasonBench",
+      "source_data": {
+        "dataset_name": "ReasonBench",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/cistine/ReasonBench"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy on tasks evaluating stylistic reasoning, such as additive/subtractive changes, symmetry, and black & white operations. This tests a VLM's sensitivity to graphical transformations and visual modifications.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Stylistic Reasoning Accuracy on ReasonBench"
+        },
+        "metric_id": "stylistic_reasoning_accuracy_on_reasonbench",
+        "metric_name": "Stylistic Reasoning Accuracy on ReasonBench",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 31.2
+      },
+      "evaluation_result_id": "ReasonBench/InternVL2/1771591481.616601#reasonbench#stylistic_reasoning_accuracy_on_reasonbench"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/42/0042af3a-8a3f-4ffc-9cfe-939be5708cf4.json

@@ -0,0 +1,58 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "LiveCodeBench/CodeQwen15-7B-Chat/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "UC Berkeley",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "CodeQwen15-7B-Chat",
+    "name": "CodeQwen15-7B-Chat",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "LiveCodeBench",
+      "source_data": {
+        "dataset_name": "LiveCodeBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2403.07974"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall Pass@1 scores for the code generation task, as reported in the original paper. This evaluation uses a contamination-free subset of problems released after September 2023. The task assesses an LLM's ability to translate natural language descriptions into functional Python code.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (%) - Code Generation (Paper)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "LiveCodeBench: Code Generation Performance (Paper Results)"
+        },
+        "metric_id": "livecodebench_code_generation_performance_paper_results",
+        "metric_name": "LiveCodeBench: Code Generation Performance (Paper Results)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 17.6
+      },
+      "evaluation_result_id": "LiveCodeBench/CodeQwen15-7B-Chat/1771591481.616601#livecodebench#livecodebench_code_generation_performance_paper_results"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/45/0045a835-025c-4c25-86aa-4f5d3be4ffda.json

@@ -0,0 +1,268 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "MIHBench/InternVL2.5/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Xiamen University",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "InternVL2.5",
+    "name": "InternVL2.5",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "MIHBench",
+      "source_data": {
+        "dataset_name": "MIHBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.00726"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the multi-image object existence hallucination task of MIHBench. This task assesses an MLLM's ability to accurately determine if a specific object exists in all provided images. Results are averaged over random, popular, and adversarial subsets. Higher is better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "MIHBench Existence Task - Accuracy"
+        },
+        "metric_id": "mihbench_existence_task_accuracy",
+        "metric_name": "MIHBench Existence Task - Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 74
+      },
+      "evaluation_result_id": "MIHBench/InternVL2.5/1771591481.616601#mihbench#mihbench_existence_task_accuracy"
+    },
+    {
+      "evaluation_name": "MIHBench",
+      "source_data": {
+        "dataset_name": "MIHBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.00726"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "F1 Score on the multi-image object count hallucination task of MIHBench. The low scores across models indicate this is a particularly challenging task, requiring fine-grained quantitative reasoning. Higher is better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1 Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "MIHBench Count Task - F1 Score"
+        },
+        "metric_id": "mihbench_count_task_f1_score",
+        "metric_name": "MIHBench Count Task - F1 Score",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 16.48
+      },
+      "evaluation_result_id": "MIHBench/InternVL2.5/1771591481.616601#mihbench#mihbench_count_task_f1_score"
+    },
+    {
+      "evaluation_name": "MIHBench",
+      "source_data": {
+        "dataset_name": "MIHBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.00726"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "F1 Score on the multi-image object existence hallucination task of MIHBench. The F1 score is the harmonic mean of precision and recall, providing a balanced measure of a model's performance on this binary classification task. Higher is better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1 Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "MIHBench Existence Task - F1 Score"
+        },
+        "metric_id": "mihbench_existence_task_f1_score",
+        "metric_name": "MIHBench Existence Task - F1 Score",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 69.16
+      },
+      "evaluation_result_id": "MIHBench/InternVL2.5/1771591481.616601#mihbench#mihbench_existence_task_f1_score"
+    },
+    {
+      "evaluation_name": "MIHBench",
+      "source_data": {
+        "dataset_name": "MIHBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.00726"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The proportion of 'Yes' responses generated by the model on the multi-image object existence task. As the dataset is balanced between positive and negative samples, a YES Ratio close to 50% indicates less bias. High deviations can indicate a tendency to hallucinate.",
+        "additional_details": {
+          "alphaxiv_y_axis": "YES Ratio (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "MIHBench Existence Task - YES Ratio"
+        },
+        "metric_id": "mihbench_existence_task_yes_ratio",
+        "metric_name": "MIHBench Existence Task - YES Ratio",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 33.42
+      },
+      "evaluation_result_id": "MIHBench/InternVL2.5/1771591481.616601#mihbench#mihbench_existence_task_yes_ratio"
+    },
+    {
+      "evaluation_name": "MIHBench",
+      "source_data": {
+        "dataset_name": "MIHBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.00726"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the object identity consistency hallucination task of MIHBench. This task assesses a model's ability to maintain object identity across multiple images, especially when distractor instances are present. Higher is better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "MIHBench Identity Consistency Task - Accuracy"
+        },
+        "metric_id": "mihbench_identity_consistency_task_accuracy",
+        "metric_name": "MIHBench Identity Consistency Task - Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 71.38
+      },
+      "evaluation_result_id": "MIHBench/InternVL2.5/1771591481.616601#mihbench#mihbench_identity_consistency_task_accuracy"
+    },
+    {
+      "evaluation_name": "MIHBench",
+      "source_data": {
+        "dataset_name": "MIHBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.00726"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "F1 Score on the object identity consistency hallucination task of MIHBench, which measures a model's ability to recognize the same object instance across multiple views. Higher is better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "F1 Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "MIHBench Identity Consistency Task - F1 Score"
+        },
+        "metric_id": "mihbench_identity_consistency_task_f1_score",
+        "metric_name": "MIHBench Identity Consistency Task - F1 Score",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 74.92
+      },
+      "evaluation_result_id": "MIHBench/InternVL2.5/1771591481.616601#mihbench#mihbench_identity_consistency_task_f1_score"
+    },
+    {
+      "evaluation_name": "MIHBench",
+      "source_data": {
+        "dataset_name": "MIHBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.00726"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "The proportion of 'Yes' responses generated by models on the identity consistency task. The extremely high ratio (e.g., 97.13% for LLaVA-NeXT-Interleave) indicates a strong bias towards confirming consistency, a key sign of hallucination.",
+        "additional_details": {
+          "alphaxiv_y_axis": "YES Ratio (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "MIHBench Identity Consistency Task - YES Ratio"
+        },
+        "metric_id": "mihbench_identity_consistency_task_yes_ratio",
+        "metric_name": "MIHBench Identity Consistency Task - YES Ratio",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 64.13
+      },
+      "evaluation_result_id": "MIHBench/InternVL2.5/1771591481.616601#mihbench#mihbench_identity_consistency_task_yes_ratio"
+    },
+    {
+      "evaluation_name": "MIHBench",
+      "source_data": {
+        "dataset_name": "MIHBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2508.00726"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the multi-image object count hallucination task of MIHBench. This task evaluates a model's capacity to accurately compare the counts of a specific object category across two images. Higher is better.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "MIHBench Count Task - Accuracy"
+        },
+        "metric_id": "mihbench_count_task_accuracy",
+        "metric_name": "MIHBench Count Task - Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 53.13
+      },
+      "evaluation_result_id": "MIHBench/InternVL2.5/1771591481.616601#mihbench#mihbench_count_task_accuracy"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/45/0045baaa-1acb-488c-acb6-32f5b66ae030.json

@@ -0,0 +1,88 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "CHURRO-DS/Azure OCR/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Stanford University",
+      "alphaxiv_dataset_type": "document",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Azure OCR",
+    "name": "Azure OCR",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "CHURRO-DS",
+      "source_data": {
+        "dataset_name": "CHURRO-DS",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.19768"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average normalized Levenshtein similarity scores of models on the handwritten subset of the CHURRO-DS test set. This metric measures the character-level accuracy of transcribing historical handwritten documents. Scores range from 0 to 100, where higher is better. All models were evaluated in a zero-shot setting, except for the paper's model, CHURRO, which was fine-tuned on the dataset.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Normalized Levenshtein Similarity (Handwritten)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Historical Text Recognition Accuracy on Handwritten Documents (CHURRO-DS)"
+        },
+        "metric_id": "historical_text_recognition_accuracy_on_handwritten_documents_churro_ds",
+        "metric_name": "Historical Text Recognition Accuracy on Handwritten Documents (CHURRO-DS)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 47.7
+      },
+      "evaluation_result_id": "CHURRO-DS/Azure OCR/1771591481.616601#churro_ds#historical_text_recognition_accuracy_on_handwritten_documents_churro_ds"
+    },
+    {
+      "evaluation_name": "CHURRO-DS",
+      "source_data": {
+        "dataset_name": "CHURRO-DS",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2509.19768"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average normalized Levenshtein similarity scores of models on the printed subset of the CHURRO-DS test set. This metric measures the character-level accuracy of transcribing historical printed documents. Scores range from 0 to 100, where higher is better. All models were evaluated in a zero-shot setting, except for the paper's model, CHURRO, which was fine-tuned on the dataset.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Avg. Normalized Levenshtein Similarity (Printed)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Historical Text Recognition Accuracy on Printed Documents (CHURRO-DS)"
+        },
+        "metric_id": "historical_text_recognition_accuracy_on_printed_documents_churro_ds",
+        "metric_name": "Historical Text Recognition Accuracy on Printed Documents (CHURRO-DS)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 71.9
+      },
+      "evaluation_result_id": "CHURRO-DS/Azure OCR/1771591481.616601#churro_ds#historical_text_recognition_accuracy_on_printed_documents_churro_ds"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/49/00497636-0411-4f75-a6a8-7b2ddf8ba5ce.json

@@ -0,0 +1,298 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "SOLIDGEO/Human/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Chinese Academy of Sciences",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Human",
+    "name": "Human",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall accuracy of various Multimodal Large Language Models (MLLMs) and text-only models on the SOLIDGEO benchmark, which evaluates mathematical reasoning in solid geometry. This metric represents the percentage of correctly answered questions across all categories. Human performance is included as a baseline. The evaluation was conducted using a zero-shot direct answering setting.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Accuracy (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Overall Performance on the SOLIDGEO Benchmark"
+        },
+        "metric_id": "overall_performance_on_the_solidgeo_benchmark",
+        "metric_name": "Overall Performance on the SOLIDGEO Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 77.5
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#overall_performance_on_the_solidgeo_benchmark"
+    },
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Model performance on the Composite Solid Structures (CSS) category of the SOLIDGEO benchmark. This task involves problems with complex solids formed by combining, intersecting, or modifying standard geometric shapes.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - CSS",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "SOLIDGEO: Composite Solid Structures (CSS) Accuracy"
+        },
+        "metric_id": "solidgeo_composite_solid_structures_css_accuracy",
+        "metric_name": "SOLIDGEO: Composite Solid Structures (CSS) Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 88.2
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#solidgeo_composite_solid_structures_css_accuracy"
+    },
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Model performance on the Measurement of Solid Geometric Forms (MSGF) category of the SOLIDGEO benchmark. This task focuses on fundamental formula-based computations for standard 3D shapes like cubes, cylinders, cones, and spheres.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - MSGF",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "SOLIDGEO: Measurement of Solid Geometric Forms (MSGF) Accuracy"
+        },
+        "metric_id": "solidgeo_measurement_of_solid_geometric_forms_msgf_accuracy",
+        "metric_name": "SOLIDGEO: Measurement of Solid Geometric Forms (MSGF) Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 87.4
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#solidgeo_measurement_of_solid_geometric_forms_msgf_accuracy"
+    },
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Model performance on the Multi-view Projection (MVP) category of the SOLIDGEO benchmark. This task evaluates the ability to interpret orthographic projections and switch between 2D views (front, top, side) and 3D spatial understanding.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - MVP",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "SOLIDGEO: Multi-view Projection (MVP) Accuracy"
+        },
+        "metric_id": "solidgeo_multi_view_projection_mvp_accuracy",
+        "metric_name": "SOLIDGEO: Multi-view Projection (MVP) Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 78.5
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#solidgeo_multi_view_projection_mvp_accuracy"
+    },
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Model performance on the Planar Unfolding and Configuration (PUC) category of the SOLIDGEO benchmark. This task involves analyzing how 3D solids unfold into 2D nets and vice versa, requiring spatial folding logic and surface pathfinding.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - PUC",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "SOLIDGEO: Planar Unfolding and Configuration (PUC) Accuracy"
+        },
+        "metric_id": "solidgeo_planar_unfolding_and_configuration_puc_accuracy",
+        "metric_name": "SOLIDGEO: Planar Unfolding and Configuration (PUC) Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 77.2
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#solidgeo_planar_unfolding_and_configuration_puc_accuracy"
+    },
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Model performance on the Solid Geometry Modeling (SGM) category of the SOLIDGEO benchmark. This is an application-oriented category with problems simulating real-world use cases like optimization, design, and constraint validation.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - SGM",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "SOLIDGEO: Solid Geometry Modeling (SGM) Accuracy"
+        },
+        "metric_id": "solidgeo_solid_geometry_modeling_sgm_accuracy",
+        "metric_name": "SOLIDGEO: Solid Geometry Modeling (SGM) Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 71.2
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#solidgeo_solid_geometry_modeling_sgm_accuracy"
+    },
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Model performance on the Spatial Metric Relations (SMR) category of the SOLIDGEO benchmark. This task involves reasoning about geometric measurements in 3D space, such as distances, angles, and relative positions, often requiring the application of geometric theorems.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - SMR",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "SOLIDGEO: Spatial Metric Relations (SMR) Accuracy"
+        },
+        "metric_id": "solidgeo_spatial_metric_relations_smr_accuracy",
+        "metric_name": "SOLIDGEO: Spatial Metric Relations (SMR) Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 70.9
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#solidgeo_spatial_metric_relations_smr_accuracy"
+    },
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Model performance on the 3D Coordinate and Vector Reasoning (3DCV) category of the SOLIDGEO benchmark. This task involves using algebraic methods, such as coordinate geometry and vector calculations, to solve geometric problems in 3D space.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - 3DCV",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "SOLIDGEO: 3D Coordinate and Vector Reasoning (3DCV) Accuracy"
+        },
+        "metric_id": "solidgeo_3d_coordinate_and_vector_reasoning_3dcv_accuracy",
+        "metric_name": "SOLIDGEO: 3D Coordinate and Vector Reasoning (3DCV) Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 69.2
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#solidgeo_3d_coordinate_and_vector_reasoning_3dcv_accuracy"
+    },
+    {
+      "evaluation_name": "SOLIDGEO",
+      "source_data": {
+        "dataset_name": "SOLIDGEO",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/HarryYancy/SolidGeo/"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Model performance on the Solid Shape Identification (SSI) category of the SOLIDGEO benchmark. This task requires the recognition and naming of 3D geometric solids or their components based on visual or structural cues.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) - SSI",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "SOLIDGEO: Solid Shape Identification (SSI) Accuracy"
+        },
+        "metric_id": "solidgeo_solid_shape_identification_ssi_accuracy",
+        "metric_name": "SOLIDGEO: Solid Shape Identification (SSI) Accuracy",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 90.2
+      },
+      "evaluation_result_id": "SOLIDGEO/Human/1771591481.616601#solidgeo#solidgeo_solid_shape_identification_ssi_accuracy"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/49/0049f31c-74f7-49d5-ad20-35b5539ffd2f.json

@@ -0,0 +1,169 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "hfopenllm_v2/SaisExperiments_QwOwO-7B-V1/1773936498.240187",
+  "retrieved_timestamp": "1773936498.240187",
+  "source_metadata": {
+    "source_name": "HF Open LLM v2",
+    "source_type": "documentation",
+    "source_organization_name": "Hugging Face",
+    "evaluator_relationship": "third_party"
+  },
+  "eval_library": {
+    "name": "lm-evaluation-harness",
+    "version": "0.4.0",
+    "additional_details": {
+      "fork": "https://github.com/huggingface/lm-evaluation-harness/tree/adding_all_changess"
+    }
+  },
+  "model_info": {
+    "name": "QwOwO-7B-V1",
+    "id": "SaisExperiments/QwOwO-7B-V1",
+    "developer": "SaisExperiments",
+    "inference_platform": "unknown",
+    "additional_details": {
+      "precision": "bfloat16",
+      "architecture": "Qwen2ForCausalLM",
+      "params_billions": "7.616"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "IFEval",
+      "source_data": {
+        "dataset_name": "IFEval",
+        "source_type": "hf_dataset",
+        "hf_repo": "google/IFEval"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on IFEval",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.4556
+      },
+      "evaluation_result_id": "hfopenllm_v2/SaisExperiments_QwOwO-7B-V1/1773936498.240187#ifeval#accuracy"
+    },
+    {
+      "evaluation_name": "BBH",
+      "source_data": {
+        "dataset_name": "BBH",
+        "source_type": "hf_dataset",
+        "hf_repo": "SaylorTwift/bbh"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on BBH",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.5431
+      },
+      "evaluation_result_id": "hfopenllm_v2/SaisExperiments_QwOwO-7B-V1/1773936498.240187#bbh#accuracy"
+    },
+    {
+      "evaluation_name": "MATH Level 5",
+      "source_data": {
+        "dataset_name": "MATH Level 5",
+        "source_type": "hf_dataset",
+        "hf_repo": "DigitalLearningGmbH/MATH-lighteval"
+      },
+      "metric_config": {
+        "evaluation_description": "Exact Match on MATH Level 5",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "exact_match",
+        "metric_name": "Exact Match",
+        "metric_kind": "exact_match",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.386
+      },
+      "evaluation_result_id": "hfopenllm_v2/SaisExperiments_QwOwO-7B-V1/1773936498.240187#math_level_5#exact_match"
+    },
+    {
+      "evaluation_name": "GPQA",
+      "source_data": {
+        "dataset_name": "GPQA",
+        "source_type": "hf_dataset",
+        "hf_repo": "Idavidrein/gpqa"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on GPQA",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.2601
+      },
+      "evaluation_result_id": "hfopenllm_v2/SaisExperiments_QwOwO-7B-V1/1773936498.240187#gpqa#accuracy"
+    },
+    {
+      "evaluation_name": "MUSR",
+      "source_data": {
+        "dataset_name": "MUSR",
+        "source_type": "hf_dataset",
+        "hf_repo": "TAUR-Lab/MuSR"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on MUSR",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.3835
+      },
+      "evaluation_result_id": "hfopenllm_v2/SaisExperiments_QwOwO-7B-V1/1773936498.240187#musr#accuracy"
+    },
+    {
+      "evaluation_name": "MMLU-PRO",
+      "source_data": {
+        "dataset_name": "MMLU-PRO",
+        "source_type": "hf_dataset",
+        "hf_repo": "TIGER-Lab/MMLU-Pro"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on MMLU-PRO",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.4224
+      },
+      "evaluation_result_id": "hfopenllm_v2/SaisExperiments_QwOwO-7B-V1/1773936498.240187#mmlu_pro#accuracy"
+    }
+  ]
+}

flat/objects/00/4a/004ad6a4-2696-449f-99f7-d46f8b30b777.json

@@ -0,0 +1,358 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "D-Judge/Versatile Diffusion/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Sun Yat-Sen University",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Versatile Diffusion",
+    "name": "Versatile Diffusion",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Image Recognition Discrepancy Rate (DR_re) for AI-generated images under Text-to-Image (T2I) guidance. This evaluates their applicability in a downstream image recognition task using a pre-trained ResNet-152 model. Lower scores indicate less discrepancy with natural images and thus better performance in this context.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Image Recognition Discrepancy Rate (%) - T2I",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Image Recognition Discrepancy Rate on D-Judge (T2I)"
+        },
+        "metric_id": "image_recognition_discrepancy_rate_on_d_judge_t2i",
+        "metric_name": "Image Recognition Discrepancy Rate on D-Judge (T2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 68.12
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#image_recognition_discrepancy_rate_on_d_judge_t2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average human perception score (on a 1-5 scale, higher is better) for semantic alignment of AI-generated images under Text-to-Image (T2I) guidance. This evaluates how well the generated images align with the text prompt from a human perspective.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Human Score (Alignment, 1-5) - T2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Human-Rated Semantic Alignment on D-Judge (T2I)"
+        },
+        "metric_id": "human_rated_semantic_alignment_on_d_judge_t2i",
+        "metric_name": "Human-Rated Semantic Alignment on D-Judge (T2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 4.22
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#human_rated_semantic_alignment_on_d_judge_t2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average human perception score (on a 1-5 scale, higher is better) for AI-generated image quality under Text-to-Image (T2I) guidance. This evaluates how closely the generated images' quality matches human expectations.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Human Score (Quality, 1-5) - T2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Human-Rated Image Quality on D-Judge (T2I)"
+        },
+        "metric_id": "human_rated_image_quality_on_d_judge_t2i",
+        "metric_name": "Human-Rated Image Quality on D-Judge (T2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 4.11
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#human_rated_image_quality_on_d_judge_t2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Image Recognition Discrepancy Rate (DR_re) for AI-generated images under Image-to-Image (I2I) guidance. This evaluates their applicability in a downstream image recognition task using a pre-trained ResNet-152 model. Lower scores indicate less discrepancy with natural images and thus better performance in this context.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Image Recognition Discrepancy Rate (%) - I2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Image Recognition Discrepancy Rate on D-Judge (I2I)"
+        },
+        "metric_id": "image_recognition_discrepancy_rate_on_d_judge_i2i",
+        "metric_name": "Image Recognition Discrepancy Rate on D-Judge (I2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 40.9
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#image_recognition_discrepancy_rate_on_d_judge_i2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Image Recognition Discrepancy Rate (DR_re) for AI-generated images under Text-and-Image-to-Image (TI2I) guidance. This evaluates their applicability in a downstream image recognition task using a pre-trained ResNet-152 model. Lower scores indicate less discrepancy with natural images and thus better performance in this context.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Image Recognition Discrepancy Rate (%) - TI2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Image Recognition Discrepancy Rate on D-Judge (TI2I)"
+        },
+        "metric_id": "image_recognition_discrepancy_rate_on_d_judge_ti2i",
+        "metric_name": "Image Recognition Discrepancy Rate on D-Judge (TI2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 60.05
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#image_recognition_discrepancy_rate_on_d_judge_ti2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Intersection over Union (IoU) scores for object segmentation performed on AI-generated images under Image-to-Image (I2I) guidance, using U2Net. Higher scores indicate better segmentation accuracy and applicability for this downstream task.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Object Segmentation IoU - I2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Object Segmentation IoU on D-Judge (I2I)"
+        },
+        "metric_id": "object_segmentation_iou_on_d_judge_i2i",
+        "metric_name": "Object Segmentation IoU on D-Judge (I2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.69
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#object_segmentation_iou_on_d_judge_i2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Intersection over Union (IoU) scores for object segmentation performed on AI-generated images under Text-to-Image (T2I) guidance, using U2Net. Higher scores indicate better segmentation accuracy and applicability for this downstream task.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Object Segmentation IoU - T2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Object Segmentation IoU on D-Judge (T2I)"
+        },
+        "metric_id": "object_segmentation_iou_on_d_judge_t2i",
+        "metric_name": "Object Segmentation IoU on D-Judge (T2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.26
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#object_segmentation_iou_on_d_judge_t2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Intersection over Union (IoU) scores for object segmentation performed on AI-generated images under Text-and-Image-to-Image (TI2I) guidance, using U2Net. Higher scores indicate better segmentation accuracy and applicability for this downstream task.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Object Segmentation IoU - TI2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Object Segmentation IoU on D-Judge (TI2I)"
+        },
+        "metric_id": "object_segmentation_iou_on_d_judge_ti2i",
+        "metric_name": "Object Segmentation IoU on D-Judge (TI2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0.29
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#object_segmentation_iou_on_d_judge_ti2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Difference Rate of CLIP Scores (DR_CS) for models under Text-to-Image (T2I) guidance. It assesses the semantic alignment discrepancy between generated images and their text prompts. Lower scores indicate less discrepancy and better semantic alignment.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Semantic Alignment Discrepancy (%) - T2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Semantic Alignment Discrepancy on D-Judge (T2I)"
+        },
+        "metric_id": "semantic_alignment_discrepancy_on_d_judge_t2i",
+        "metric_name": "Semantic Alignment Discrepancy on D-Judge (T2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 9.71
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#semantic_alignment_discrepancy_on_d_judge_t2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average human perception score (on a 1-5 scale, higher is better) for the aesthetic appeal of AI-generated images under Text-to-Image (T2I) guidance. This evaluates the artistic quality of the images as judged by humans.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Human Score (Aesthetic, 1-5) - T2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Human-Rated Aesthetic Appeal on D-Judge (T2I)"
+        },
+        "metric_id": "human_rated_aesthetic_appeal_on_d_judge_t2i",
+        "metric_name": "Human-Rated Aesthetic Appeal on D-Judge (T2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 3.96
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#human_rated_aesthetic_appeal_on_d_judge_t2i"
+    },
+    {
+      "evaluation_name": "D-Judge",
+      "source_data": {
+        "dataset_name": "D-Judge",
+        "source_type": "url",
+        "url": [
+          "https://huggingface.co/datasets/Renyang/DANI"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Difference Rate (DR) for the Visual Question Answering (VQA) task of object identification on images generated with Text-to-Image (T2I) guidance. This measures the model's ability to generate images with clearly identifiable objects. Lower scores indicate less discrepancy with natural images.",
+        "additional_details": {
+          "alphaxiv_y_axis": "VQA Object Discrepancy Rate (%) - T2I",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "VQA Object Identification Discrepancy on D-Judge (T2I)"
+        },
+        "metric_id": "vqa_object_identification_discrepancy_on_d_judge_t2i",
+        "metric_name": "VQA Object Identification Discrepancy on D-Judge (T2I)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 74.21
+      },
+      "evaluation_result_id": "D-Judge/Versatile Diffusion/1771591481.616601#d_judge#vqa_object_identification_discrepancy_on_d_judge_t2i"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/50/005076b3-3faf-44b3-afa4-8c0ecfd4d23f.json

@@ -0,0 +1,113 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "sciarena/google/gemini-3-pro-preview/1775549757.5677671",
+  "retrieved_timestamp": "1775549757.5677671",
+  "source_metadata": {
+    "source_name": "SciArena leaderboard API",
+    "source_type": "documentation",
+    "source_organization_name": "Ai2",
+    "source_organization_url": "https://sciarena.allen.ai",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "api_endpoint": "https://sciarena.allen.ai/api/leaderboard"
+    }
+  },
+  "eval_library": {
+    "name": "SciArena",
+    "version": "unknown"
+  },
+  "model_info": {
+    "name": "Gemini-3-Pro-Preview",
+    "id": "google/gemini-3-pro-preview",
+    "developer": "google",
+    "additional_details": {
+      "raw_model_id": "Gemini-3-Pro-Preview"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_result_id": "sciarena/google/gemini-3-pro-preview/1775549757.5677671#overall_elo#elo",
+      "evaluation_name": "overall_elo",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "SciArena leaderboard API",
+        "url": [
+          "https://sciarena.allen.ai/api/leaderboard"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "elo",
+        "metric_name": "Elo rating",
+        "metric_type": "continuous",
+        "metric_kind": "elo",
+        "metric_unit": "points",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 829.7737302958208,
+        "max_score": 1151.3779287263492,
+        "additional_details": {
+          "raw_metric_field": "rating"
+        }
+      },
+      "score_details": {
+        "score": 1086.3325465626408,
+        "details": {
+          "num_battles": "215",
+          "rating_q025": "1051.8638049003746",
+          "rating_q975": "1133.4082199103375",
+          "variance": "None"
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "sciarena/google/gemini-3-pro-preview/1775549757.5677671#overall_rank#rank",
+      "evaluation_name": "overall_rank",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "SciArena leaderboard API",
+        "url": [
+          "https://sciarena.allen.ai/api/leaderboard"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "rank",
+        "metric_name": "Rank",
+        "metric_type": "continuous",
+        "metric_kind": "rank",
+        "metric_unit": "position",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 1.0,
+        "max_score": 38.0
+      },
+      "score_details": {
+        "score": 4.0
+      }
+    },
+    {
+      "evaluation_result_id": "sciarena/google/gemini-3-pro-preview/1775549757.5677671#overall_cost_per_100_calls_usd#cost_per_100_calls_usd",
+      "evaluation_name": "overall_cost_per_100_calls_usd",
+      "source_data": {
+        "source_type": "url",
+        "dataset_name": "SciArena leaderboard API",
+        "url": [
+          "https://sciarena.allen.ai/api/leaderboard"
+        ]
+      },
+      "metric_config": {
+        "metric_id": "cost_per_100_calls_usd",
+        "metric_name": "Cost per 100 calls",
+        "metric_type": "continuous",
+        "metric_kind": "cost",
+        "metric_unit": "usd",
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 28.648
+      },
+      "score_details": {
+        "score": 3.3879
+      }
+    }
+  ]
+}

flat/objects/00/51/005130d2-64cc-44af-8eef-1d424a5b63f4.json

@@ -0,0 +1,208 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "CodeEval-Pro/Yi-Coder-9B-chat/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Tsinghua University",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Yi-Coder-9B-chat",
+    "name": "Yi-Coder-9B-chat",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "CodeEval-Pro",
+      "source_data": {
+        "dataset_name": "CodeEval-Pro",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2412.21199"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@1 score of models on the self-invoking problems of the HumanEval Pro benchmark. This task evaluates a model's ability to solve a complex problem that requires calling a function it generated for a simpler, related 'base problem'. This tests multi-step reasoning and code utilization. Results are from the official project leaderboard using a greedy generation strategy.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@1 (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Code Generation on HumanEval Pro (Self-invoking)"
+        },
+        "metric_id": "code_generation_on_humaneval_pro_self_invoking",
+        "metric_name": "Code Generation on HumanEval Pro (Self-invoking)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 59.8
+      },
+      "evaluation_result_id": "CodeEval-Pro/Yi-Coder-9B-chat/1771591481.616601#codeeval_pro#code_generation_on_humaneval_pro_self_invoking"
+    },
+    {
+      "evaluation_name": "CodeEval-Pro",
+      "source_data": {
+        "dataset_name": "CodeEval-Pro",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2412.21199"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@10 score on the HumanEval Pro benchmark. This metric evaluates the percentage of problems for which at least one of ten generated samples passes all unit tests, using a random sampling strategy (temperature=0.2, top_p=0.95).",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@10 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Code Generation on HumanEval Pro (pass@10)"
+        },
+        "metric_id": "code_generation_on_humaneval_pro_pass_10",
+        "metric_name": "Code Generation on HumanEval Pro (pass@10)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 67.9
+      },
+      "evaluation_result_id": "CodeEval-Pro/Yi-Coder-9B-chat/1771591481.616601#codeeval_pro#code_generation_on_humaneval_pro_pass_10"
+    },
+    {
+      "evaluation_name": "CodeEval-Pro",
+      "source_data": {
+        "dataset_name": "CodeEval-Pro",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2412.21199"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@5 score on the HumanEval Pro benchmark. This metric evaluates the percentage of problems for which at least one of five generated samples passes all unit tests, using a random sampling strategy (temperature=0.2, top_p=0.95).",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@5 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Code Generation on HumanEval Pro (pass@5)"
+        },
+        "metric_id": "code_generation_on_humaneval_pro_pass_5",
+        "metric_name": "Code Generation on HumanEval Pro (pass@5)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 66.4
+      },
+      "evaluation_result_id": "CodeEval-Pro/Yi-Coder-9B-chat/1771591481.616601#codeeval_pro#code_generation_on_humaneval_pro_pass_5"
+    },
+    {
+      "evaluation_name": "CodeEval-Pro",
+      "source_data": {
+        "dataset_name": "CodeEval-Pro",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2412.21199"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@10 score on the MBPP Pro benchmark. This metric evaluates the percentage of problems for which at least one of ten generated samples passes all unit tests, using a random sampling strategy (temperature=0.2, top_p=0.95).",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@10 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Code Generation on MBPP Pro (pass@10)"
+        },
+        "metric_id": "code_generation_on_mbpp_pro_pass_10",
+        "metric_name": "Code Generation on MBPP Pro (pass@10)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 71.2
+      },
+      "evaluation_result_id": "CodeEval-Pro/Yi-Coder-9B-chat/1771591481.616601#codeeval_pro#code_generation_on_mbpp_pro_pass_10"
+    },
+    {
+      "evaluation_name": "CodeEval-Pro",
+      "source_data": {
+        "dataset_name": "CodeEval-Pro",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2412.21199"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@5 score on the MBPP Pro benchmark. This metric evaluates the percentage of problems for which at least one of five generated samples passes all unit tests, using a random sampling strategy (temperature=0.2, top_p=0.95).",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@5 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Code Generation on MBPP Pro (pass@5)"
+        },
+        "metric_id": "code_generation_on_mbpp_pro_pass_5",
+        "metric_name": "Code Generation on MBPP Pro (pass@5)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 69.8
+      },
+      "evaluation_result_id": "CodeEval-Pro/Yi-Coder-9B-chat/1771591481.616601#codeeval_pro#code_generation_on_mbpp_pro_pass_5"
+    },
+    {
+      "evaluation_name": "CodeEval-Pro",
+      "source_data": {
+        "dataset_name": "CodeEval-Pro",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2412.21199"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@1 score of models on the self-invoking problems of the MBPP Pro benchmark. This task evaluates a model's ability to solve a complex problem that requires calling a function it generated for a simpler, related 'base problem'. This tests multi-step reasoning and code utilization. Results are from the official project leaderboard using a greedy generation strategy.",
+        "additional_details": {
+          "alphaxiv_y_axis": "pass@1 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Code Generation on MBPP Pro (Self-invoking)"
+        },
+        "metric_id": "code_generation_on_mbpp_pro_self_invoking",
+        "metric_name": "Code Generation on MBPP Pro (Self-invoking)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 64.8
+      },
+      "evaluation_result_id": "CodeEval-Pro/Yi-Coder-9B-chat/1771591481.616601#codeeval_pro#code_generation_on_mbpp_pro_self_invoking"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/51/005162a6-b099-4eca-8aaf-28adb8f11607.json

@@ -0,0 +1,208 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "CodeCrash/Qwen2.5-32B-Instruct (CoT)/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "The Chinese University of Hong Kong",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Qwen2.5-32B-Instruct (CoT)",
+    "name": "Qwen2.5-32B-Instruct (CoT)",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "CodeCrash",
+      "source_data": {
+        "dataset_name": "CodeCrash",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2504.14119"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average Pass@1 accuracy across all four perturbation types on the CODECRASH benchmark (PSC-ALL, MCC, MPS, MHC), aggregated over the CRUX and LCB datasets. This metric provides a holistic measure of a model's robustness and reliability in code reasoning when faced with various forms of misleading structural and natural language information.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (%) - Average",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "CodeCrash: Average Robustness to Perturbations"
+        },
+        "metric_id": "codecrash_average_robustness_to_perturbations",
+        "metric_name": "CodeCrash: Average Robustness to Perturbations",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 66.4
+      },
+      "evaluation_result_id": "CodeCrash/Qwen2.5-32B-Instruct (CoT)/1771591481.616601#codecrash#codecrash_average_robustness_to_perturbations"
+    },
+    {
+      "evaluation_name": "CodeCrash",
+      "source_data": {
+        "dataset_name": "CodeCrash",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2504.14119"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Pass@1 accuracy under Misleading Code Comments (MCC) perturbation, where comments that contradict the code's logic are inserted. This metric assesses a model's ability to prioritize executable code semantics over distracting and incorrect natural language information.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (%) - Misleading Comments (MCC)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "CodeCrash: Robustness to Misleading Code Comments (MCC)"
+        },
+        "metric_id": "codecrash_robustness_to_misleading_code_comments_mcc",
+        "metric_name": "CodeCrash: Robustness to Misleading Code Comments (MCC)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 64.7
+      },
+      "evaluation_result_id": "CodeCrash/Qwen2.5-32B-Instruct (CoT)/1771591481.616601#codecrash#codecrash_robustness_to_misleading_code_comments_mcc"
+    },
+    {
+      "evaluation_name": "CodeCrash",
+      "source_data": {
+        "dataset_name": "CodeCrash",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2504.14119"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Pass@1 accuracy under Misleading Hint Comments (MHC) perturbation, where plausible but incorrect high-level hints about the program's output are added as comments. This metric stress-tests a model's critical reasoning and its ability to avoid 'rationalization'—producing faulty logic to align with an incorrect hint.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (%) - Misleading Hint Comments (MHC)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "CodeCrash: Robustness to Misleading Hint Comments (MHC)"
+        },
+        "metric_id": "codecrash_robustness_to_misleading_hint_comments_mhc",
+        "metric_name": "CodeCrash: Robustness to Misleading Hint Comments (MHC)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 75.2
+      },
+      "evaluation_result_id": "CodeCrash/Qwen2.5-32B-Instruct (CoT)/1771591481.616601#codecrash#codecrash_robustness_to_misleading_hint_comments_mhc"
+    },
+    {
+      "evaluation_name": "CodeCrash",
+      "source_data": {
+        "dataset_name": "CodeCrash",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2504.14119"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Pass@1 accuracy under Misleading Print Statements (MPS) perturbation, which embeds print statements conveying incorrect information about the code's behavior. This metric measures a model's ability to distinguish executable logic from non-functional but misleading textual output within the code.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (%) - Misleading Print Statements (MPS)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "CodeCrash: Robustness to Misleading Print Statements (MPS)"
+        },
+        "metric_id": "codecrash_robustness_to_misleading_print_statements_mps",
+        "metric_name": "CodeCrash: Robustness to Misleading Print Statements (MPS)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 60.2
+      },
+      "evaluation_result_id": "CodeCrash/Qwen2.5-32B-Instruct (CoT)/1771591481.616601#codecrash#codecrash_robustness_to_misleading_print_statements_mps"
+    },
+    {
+      "evaluation_name": "CodeCrash",
+      "source_data": {
+        "dataset_name": "CodeCrash",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2504.14119"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Pass@1 accuracy under Aggregated Structural Perturbation (PSC-ALL), which combines variable renaming, code reformatting, and garbage code insertion. This metric tests a model's ability to reason about code logic independent of its superficial syntactic structure and formatting.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (%) - Structural Perturbation (PSC-ALL)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "CodeCrash: Robustness to Structural Perturbations (PSC-ALL)"
+        },
+        "metric_id": "codecrash_robustness_to_structural_perturbations_psc_all",
+        "metric_name": "CodeCrash: Robustness to Structural Perturbations (PSC-ALL)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 65.6
+      },
+      "evaluation_result_id": "CodeCrash/Qwen2.5-32B-Instruct (CoT)/1771591481.616601#codecrash#codecrash_robustness_to_structural_perturbations_psc_all"
+    },
+    {
+      "evaluation_name": "CodeCrash",
+      "source_data": {
+        "dataset_name": "CodeCrash",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2504.14119"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Pass@1 accuracy on the vanilla (unperturbed) version of the CODECRASH benchmark, aggregated over the CRUX and LCB datasets. This score represents the baseline code reasoning capability of each model before being subjected to misleading structural or natural language perturbations.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (%) - Vanilla",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "CodeCrash: Baseline Performance (Vanilla)"
+        },
+        "metric_id": "codecrash_baseline_performance_vanilla",
+        "metric_name": "CodeCrash: Baseline Performance (Vanilla)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 79.4
+      },
+      "evaluation_result_id": "CodeCrash/Qwen2.5-32B-Instruct (CoT)/1771591481.616601#codecrash#codecrash_baseline_performance_vanilla"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/54/0054047d-ee2c-4eb3-afea-5e743b295148.json

@@ -0,0 +1,169 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "hfopenllm_v2/allura-org_MoE-Girl-1BA-7BT/1773936498.240187",
+  "retrieved_timestamp": "1773936498.240187",
+  "source_metadata": {
+    "source_name": "HF Open LLM v2",
+    "source_type": "documentation",
+    "source_organization_name": "Hugging Face",
+    "evaluator_relationship": "third_party"
+  },
+  "eval_library": {
+    "name": "lm-evaluation-harness",
+    "version": "0.4.0",
+    "additional_details": {
+      "fork": "https://github.com/huggingface/lm-evaluation-harness/tree/adding_all_changess"
+    }
+  },
+  "model_info": {
+    "name": "MoE-Girl-1BA-7BT",
+    "id": "allura-org/MoE-Girl-1BA-7BT",
+    "developer": "allura-org",
+    "inference_platform": "unknown",
+    "additional_details": {
+      "precision": "bfloat16",
+      "architecture": "OlmoeForCausalLM",
+      "params_billions": "6.919"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "IFEval",
+      "source_data": {
+        "dataset_name": "IFEval",
+        "source_type": "hf_dataset",
+        "hf_repo": "google/IFEval"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on IFEval",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.2705
+      },
+      "evaluation_result_id": "hfopenllm_v2/allura-org_MoE-Girl-1BA-7BT/1773936498.240187#ifeval#accuracy"
+    },
+    {
+      "evaluation_name": "BBH",
+      "source_data": {
+        "dataset_name": "BBH",
+        "source_type": "hf_dataset",
+        "hf_repo": "SaylorTwift/bbh"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on BBH",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.3139
+      },
+      "evaluation_result_id": "hfopenllm_v2/allura-org_MoE-Girl-1BA-7BT/1773936498.240187#bbh#accuracy"
+    },
+    {
+      "evaluation_name": "MATH Level 5",
+      "source_data": {
+        "dataset_name": "MATH Level 5",
+        "source_type": "hf_dataset",
+        "hf_repo": "DigitalLearningGmbH/MATH-lighteval"
+      },
+      "metric_config": {
+        "evaluation_description": "Exact Match on MATH Level 5",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "exact_match",
+        "metric_name": "Exact Match",
+        "metric_kind": "exact_match",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.0151
+      },
+      "evaluation_result_id": "hfopenllm_v2/allura-org_MoE-Girl-1BA-7BT/1773936498.240187#math_level_5#exact_match"
+    },
+    {
+      "evaluation_name": "GPQA",
+      "source_data": {
+        "dataset_name": "GPQA",
+        "source_type": "hf_dataset",
+        "hf_repo": "Idavidrein/gpqa"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on GPQA",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.2584
+      },
+      "evaluation_result_id": "hfopenllm_v2/allura-org_MoE-Girl-1BA-7BT/1773936498.240187#gpqa#accuracy"
+    },
+    {
+      "evaluation_name": "MUSR",
+      "source_data": {
+        "dataset_name": "MUSR",
+        "source_type": "hf_dataset",
+        "hf_repo": "TAUR-Lab/MuSR"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on MUSR",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.3436
+      },
+      "evaluation_result_id": "hfopenllm_v2/allura-org_MoE-Girl-1BA-7BT/1773936498.240187#musr#accuracy"
+    },
+    {
+      "evaluation_name": "MMLU-PRO",
+      "source_data": {
+        "dataset_name": "MMLU-PRO",
+        "source_type": "hf_dataset",
+        "hf_repo": "TIGER-Lab/MMLU-Pro"
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy on MMLU-PRO",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 1.0,
+        "metric_id": "accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "proportion"
+      },
+      "score_details": {
+        "score": 0.1218
+      },
+      "evaluation_result_id": "hfopenllm_v2/allura-org_MoE-Girl-1BA-7BT/1773936498.240187#mmlu_pro#accuracy"
+    }
+  ]
+}

flat/objects/00/54/00540c71-2c66-4300-a1b1-561d3681b823.json

@@ -0,0 +1,298 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "ProBench/Codestral-22B-v0.1/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Tianjin University",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Codestral-22B-v0.1",
+    "name": "Codestral-22B-v0.1",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of problems for which at least one of 1 generated solutions successfully passes all hidden test cases on the original competition platform. This is the primary metric for the ProBench benchmark, used to rank models.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "ProBench Pass@1 Score for Competitive Programming"
+        },
+        "metric_id": "probench_pass_1_score_for_competitive_programming",
+        "metric_name": "ProBench Pass@1 Score for Competitive Programming",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 5.08
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_1_score_for_competitive_programming"
+    },
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@1 score specifically on problems with descriptions written in Chinese. This is used to assess multilingual competence in comparison to performance on English problems.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (Chinese Problems) (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ProBench Pass@1 Score on Chinese Problems"
+        },
+        "metric_id": "probench_pass_1_score_on_chinese_problems",
+        "metric_name": "ProBench Pass@1 Score on Chinese Problems",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 7.01
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_1_score_on_chinese_problems"
+    },
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@1 score on problems categorized as 'easy'. This assesses model performance on less complex competitive programming tasks. Performance is significantly higher here compared to medium or hard problems.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (Easy Problems) (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ProBench Pass@1 Score on Easy Problems"
+        },
+        "metric_id": "probench_pass_1_score_on_easy_problems",
+        "metric_name": "ProBench Pass@1 Score on Easy Problems",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 10.7
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_1_score_on_easy_problems"
+    },
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@1 score specifically on problems with descriptions written in English. This is used to assess multilingual competence in comparison to performance on Chinese problems.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (English Problems) (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ProBench Pass@1 Score on English Problems"
+        },
+        "metric_id": "probench_pass_1_score_on_english_problems",
+        "metric_name": "ProBench Pass@1 Score on English Problems",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 3.59
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_1_score_on_english_problems"
+    },
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@1 score on problems categorized as 'hard'. This assesses model performance on highly complex competitive programming tasks. All evaluated models scored 0%, highlighting the extreme difficulty of these problems for current LLMs.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (Hard Problems) (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ProBench Pass@1 Score on Hard Problems"
+        },
+        "metric_id": "probench_pass_1_score_on_hard_problems",
+        "metric_name": "ProBench Pass@1 Score on Hard Problems",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_1_score_on_hard_problems"
+    },
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the pass@1 score on problems categorized as 'medium'. This assesses model performance on moderately complex competitive programming tasks. Performance drops dramatically from easy problems, with most models scoring below 1%.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@1 (Medium Problems) (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ProBench Pass@1 Score on Medium Problems"
+        },
+        "metric_id": "probench_pass_1_score_on_medium_problems",
+        "metric_name": "ProBench Pass@1 Score on Medium Problems",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_1_score_on_medium_problems"
+    },
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of problems for which at least one of 2 generated solutions successfully passes all hidden test cases on the original competition platform.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@2 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ProBench Pass@2 Score for Competitive Programming"
+        },
+        "metric_id": "probench_pass_2_score_for_competitive_programming",
+        "metric_name": "ProBench Pass@2 Score for Competitive Programming",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 7.08
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_2_score_for_competitive_programming"
+    },
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of problems for which at least one of 4 generated solutions successfully passes all hidden test cases on the original competition platform.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@4 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ProBench Pass@4 Score for Competitive Programming"
+        },
+        "metric_id": "probench_pass_4_score_for_competitive_programming",
+        "metric_name": "ProBench Pass@4 Score for Competitive Programming",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 9.36
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_4_score_for_competitive_programming"
+    },
+    {
+      "evaluation_name": "ProBench",
+      "source_data": {
+        "dataset_name": "ProBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.20868"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of problems for which at least one of 8 generated solutions successfully passes all hidden test cases on the original competition platform.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Pass@8 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "ProBench Pass@8 Score for Competitive Programming"
+        },
+        "metric_id": "probench_pass_8_score_for_competitive_programming",
+        "metric_name": "ProBench Pass@8 Score for Competitive Programming",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 11.65
+      },
+      "evaluation_result_id": "ProBench/Codestral-22B-v0.1/1771591481.616601#probench#probench_pass_8_score_for_competitive_programming"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/56/00560c98-24a9-42b6-acde-37b89a224587.json

@@ -0,0 +1,840 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "vals-ai/mgsm/xai_grok-4-1-fast-reasoning/1777395187.3170502",
+  "retrieved_timestamp": "1777395187.3170502",
+  "source_metadata": {
+    "source_name": "Vals.ai Leaderboard - MGSM",
+    "source_type": "documentation",
+    "source_organization_name": "Vals.ai",
+    "source_organization_url": "https://www.vals.ai",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "benchmark_slug": "mgsm",
+      "benchmark_name": "MGSM",
+      "benchmark_updated": "2026-01-09",
+      "dataset_type": "public",
+      "industry": "math",
+      "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm",
+      "extraction_method": "static_astro_benchmark_view_props"
+    }
+  },
+  "eval_library": {
+    "name": "Vals.ai",
+    "version": "unknown"
+  },
+  "model_info": {
+    "name": "grok-4-1-fast-reasoning",
+    "id": "xai/grok-4-1-fast-reasoning",
+    "developer": "xai",
+    "additional_details": {
+      "vals_model_id": "grok/grok-4-1-fast-reasoning",
+      "vals_provider": "xAI"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_result_id": "mgsm:mgsm_bn:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_bn",
+      "source_data": {
+        "dataset_name": "MGSM - Bengali",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_bn",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Bengali).",
+        "metric_id": "vals_ai.mgsm.mgsm_bn.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 87.6,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_bn",
+          "task_name": "Bengali",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "87.6",
+          "raw_stderr": "2.084",
+          "latency": "20.59",
+          "cost_per_test": "0.000735",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.084,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_de:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_de",
+      "source_data": {
+        "dataset_name": "MGSM - German",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_de",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (German).",
+        "metric_id": "vals_ai.mgsm.mgsm_de.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 90.4,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_de",
+          "task_name": "German",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "90.4",
+          "raw_stderr": "1.863",
+          "latency": "11.03",
+          "cost_per_test": "0.000448",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.863,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_en:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_en",
+      "source_data": {
+        "dataset_name": "MGSM - English",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_en",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (English).",
+        "metric_id": "vals_ai.mgsm.mgsm_en.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 95.2,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_en",
+          "task_name": "English",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "95.2",
+          "raw_stderr": "1.352",
+          "latency": "6.779",
+          "cost_per_test": "0.000298",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.352,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_es:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_es",
+      "source_data": {
+        "dataset_name": "MGSM - Spanish",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_es",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Spanish).",
+        "metric_id": "vals_ai.mgsm.mgsm_es.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 86.0,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_es",
+          "task_name": "Spanish",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "86",
+          "raw_stderr": "2.194",
+          "latency": "9.599",
+          "cost_per_test": "0.00039",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.194,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_fr:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_fr",
+      "source_data": {
+        "dataset_name": "MGSM - French",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_fr",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (French).",
+        "metric_id": "vals_ai.mgsm.mgsm_fr.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 82.0,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_fr",
+          "task_name": "French",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "82",
+          "raw_stderr": "2.43",
+          "latency": "11.581",
+          "cost_per_test": "0.000474",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.43,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_ja:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_ja",
+      "source_data": {
+        "dataset_name": "MGSM - Japanese",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_ja",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Japanese).",
+        "metric_id": "vals_ai.mgsm.mgsm_ja.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 87.2,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_ja",
+          "task_name": "Japanese",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "87.2",
+          "raw_stderr": "2.113",
+          "latency": "10.958",
+          "cost_per_test": "0.000432",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.113,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_ru:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_ru",
+      "source_data": {
+        "dataset_name": "MGSM - Russian",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_ru",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Russian).",
+        "metric_id": "vals_ai.mgsm.mgsm_ru.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 93.6,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_ru",
+          "task_name": "Russian",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "93.6",
+          "raw_stderr": "1.548",
+          "latency": "10.677",
+          "cost_per_test": "0.000401",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.548,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_sw:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_sw",
+      "source_data": {
+        "dataset_name": "MGSM - Swahili",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_sw",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Swahili).",
+        "metric_id": "vals_ai.mgsm.mgsm_sw.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 93.2,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_sw",
+          "task_name": "Swahili",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "93.2",
+          "raw_stderr": "1.592",
+          "latency": "17.427",
+          "cost_per_test": "0.00066",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.592,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_te:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_te",
+      "source_data": {
+        "dataset_name": "MGSM - Telugu",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_te",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Telugu).",
+        "metric_id": "vals_ai.mgsm.mgsm_te.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 85.2,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_te",
+          "task_name": "Telugu",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "85.2",
+          "raw_stderr": "2.246",
+          "latency": "27.775",
+          "cost_per_test": "0.001001",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.246,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_th:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_th",
+      "source_data": {
+        "dataset_name": "MGSM - Thai",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_th",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Thai).",
+        "metric_id": "vals_ai.mgsm.mgsm_th.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 93.2,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_th",
+          "task_name": "Thai",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "93.2",
+          "raw_stderr": "1.592",
+          "latency": "13.831",
+          "cost_per_test": "0.000533",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.592,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:mgsm_zh:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.mgsm_zh",
+      "source_data": {
+        "dataset_name": "MGSM - Chinese",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "mgsm_zh",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Chinese).",
+        "metric_id": "vals_ai.mgsm.mgsm_zh.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 91.2,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "mgsm_zh",
+          "task_name": "Chinese",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "91.2",
+          "raw_stderr": "1.792",
+          "latency": "8.268",
+          "cost_per_test": "0.000321",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.792,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mgsm:overall:grok/grok-4-1-fast-reasoning:score",
+      "evaluation_name": "vals_ai.mgsm.overall",
+      "source_data": {
+        "dataset_name": "MGSM - Overall",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mgsm"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mgsm",
+          "task_key": "overall",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MGSM (Overall).",
+        "metric_id": "vals_ai.mgsm.overall.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mgsm"
+        }
+      },
+      "score_details": {
+        "score": 89.527,
+        "details": {
+          "benchmark_slug": "mgsm",
+          "benchmark_name": "MGSM",
+          "benchmark_updated": "2026-01-09",
+          "task_key": "overall",
+          "task_name": "Overall",
+          "dataset_type": "public",
+          "industry": "math",
+          "raw_score": "89.527",
+          "raw_stderr": "1.892",
+          "latency": "13.501",
+          "cost_per_test": "0.000518",
+          "temperature": "0.7",
+          "top_p": "0.95",
+          "max_output_tokens": "128000",
+          "provider": "xAI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.892,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.7,
+          "top_p": 0.95,
+          "max_tokens": 128000,
+          "max_attempts": 1
+        }
+      }
+    }
+  ]
+}

flat/objects/00/59/0059205f-800e-4c1a-b2df-1354daca0783.json

@@ -0,0 +1,208 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "InteractScience/DeepSeek-V3-0324/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Shanghai Artificial Intelligence Laboratory",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "DeepSeek-V3-0324",
+    "name": "DeepSeek-V3-0324",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "InteractScience",
+      "source_data": {
+        "dataset_name": "InteractScience",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09724"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Assesses higher-level semantic features and scientific correctness of generated visualizations. Gemini-2.5-Pro acts as a Vision-Language Model (VLM) judge, guided by a checklist, to evaluate if the visual result aligns with task specifications. The raw score (0-5) is rescaled to a 0–100 range.",
+        "additional_details": {
+          "alphaxiv_y_axis": "VQT VLM-Judge Score",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Visually-Grounded Qualitative Testing (VQT) - VLM-Judge Score"
+        },
+        "metric_id": "visually_grounded_qualitative_testing_vqt_vlm_judge_score",
+        "metric_name": "Visually-Grounded Qualitative Testing (VQT) - VLM-Judge Score",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 49.46
+      },
+      "evaluation_result_id": "InteractScience/DeepSeek-V3-0324/1771591481.616601#interactscience#visually_grounded_qualitative_testing_vqt_vlm_judge_score"
+    },
+    {
+      "evaluation_name": "InteractScience",
+      "source_data": {
+        "dataset_name": "InteractScience",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09724"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the functional correctness of the generated code. It is the macro average, calculated by averaging the pass rates for each individual problem across the INTERACTSCIENCE benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PFT Average Pass Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Programmatic Functional Testing (PFT) - Average Pass Rate"
+        },
+        "metric_id": "programmatic_functional_testing_pft_average_pass_rate",
+        "metric_name": "Programmatic Functional Testing (PFT) - Average Pass Rate",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 30.57
+      },
+      "evaluation_result_id": "InteractScience/DeepSeek-V3-0324/1771591481.616601#interactscience#programmatic_functional_testing_pft_average_pass_rate"
+    },
+    {
+      "evaluation_name": "InteractScience",
+      "source_data": {
+        "dataset_name": "InteractScience",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09724"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the overall functional correctness of the generated code. It is the micro average, representing the percentage of all PFT unit tests passed across the entire INTERACTSCIENCE benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PFT Overall Pass Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Programmatic Functional Testing (PFT) - Overall Pass Rate"
+        },
+        "metric_id": "programmatic_functional_testing_pft_overall_pass_rate",
+        "metric_name": "Programmatic Functional Testing (PFT) - Overall Pass Rate",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 31.73
+      },
+      "evaluation_result_id": "InteractScience/DeepSeek-V3-0324/1771591481.616601#interactscience#programmatic_functional_testing_pft_overall_pass_rate"
+    },
+    {
+      "evaluation_name": "InteractScience",
+      "source_data": {
+        "dataset_name": "InteractScience",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09724"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the proportion of problems for which all associated PFT unit tests pass successfully. This is the strictest measure of functional correctness on the INTERACTSCIENCE benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "PFT Perfect Pass Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Programmatic Functional Testing (PFT) - Perfect Pass Rate"
+        },
+        "metric_id": "programmatic_functional_testing_pft_perfect_pass_rate",
+        "metric_name": "Programmatic Functional Testing (PFT) - Perfect Pass Rate",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 10.49
+      },
+      "evaluation_result_id": "InteractScience/DeepSeek-V3-0324/1771591481.616601#interactscience#programmatic_functional_testing_pft_perfect_pass_rate"
+    },
+    {
+      "evaluation_name": "InteractScience",
+      "source_data": {
+        "dataset_name": "InteractScience",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09724"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of test cases where a specified sequence of user actions successfully leads to a state where a snapshot can be captured. This indicates the surface-level interactivity of the generated scientific demonstration.",
+        "additional_details": {
+          "alphaxiv_y_axis": "VQT Action Success Rate (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Visually-Grounded Qualitative Testing (VQT) - Action Success Rate"
+        },
+        "metric_id": "visually_grounded_qualitative_testing_vqt_action_success_rate",
+        "metric_name": "Visually-Grounded Qualitative Testing (VQT) - Action Success Rate",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 85.93
+      },
+      "evaluation_result_id": "InteractScience/DeepSeek-V3-0324/1771591481.616601#interactscience#visually_grounded_qualitative_testing_vqt_action_success_rate"
+    },
+    {
+      "evaluation_name": "InteractScience",
+      "source_data": {
+        "dataset_name": "InteractScience",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2510.09724"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the low-level visual similarity between the generated visualization snapshot and its corresponding reference snapshot using a CLIP model. Scores are normalized to a 0–100 scale.",
+        "additional_details": {
+          "alphaxiv_y_axis": "VQT CLIP Score",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Visually-Grounded Qualitative Testing (VQT) - CLIP Score"
+        },
+        "metric_id": "visually_grounded_qualitative_testing_vqt_clip_score",
+        "metric_name": "Visually-Grounded Qualitative Testing (VQT) - CLIP Score",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 68.68
+      },
+      "evaluation_result_id": "InteractScience/DeepSeek-V3-0324/1771591481.616601#interactscience#visually_grounded_qualitative_testing_vqt_clip_score"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/59/005924a9-0f58-41df-b8f3-baaa61de9781.json

@@ -0,0 +1,1011 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "vals-ai/mmlu_pro/mistralai_mistral-large-2512/1777395187.3170502",
+  "retrieved_timestamp": "1777395187.3170502",
+  "source_metadata": {
+    "source_name": "Vals.ai Leaderboard - MMLU Pro",
+    "source_type": "documentation",
+    "source_organization_name": "Vals.ai",
+    "source_organization_url": "https://www.vals.ai",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "benchmark_slug": "mmlu_pro",
+      "benchmark_name": "MMLU Pro",
+      "benchmark_updated": "2026-04-23",
+      "dataset_type": "public",
+      "industry": "academic",
+      "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro",
+      "extraction_method": "static_astro_benchmark_view_props"
+    }
+  },
+  "eval_library": {
+    "name": "Vals.ai",
+    "version": "unknown"
+  },
+  "model_info": {
+    "name": "mistral-large-2512",
+    "id": "mistralai/mistral-large-2512",
+    "developer": "mistralai",
+    "additional_details": {
+      "vals_model_id": "mistralai/mistral-large-2512",
+      "vals_provider": "Mistral AI"
+    }
+  },
+  "evaluation_results": [
+    {
+      "evaluation_result_id": "mmlu_pro:biology:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.biology",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Biology",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "biology",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Biology).",
+        "metric_id": "vals_ai.mmlu_pro.biology.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 88.564,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "biology",
+          "task_name": "Biology",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "88.564",
+          "raw_stderr": "1.865",
+          "latency": "12.468",
+          "cost_per_test": "0.001387",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.865,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:business:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.business",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Business",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "business",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Business).",
+        "metric_id": "vals_ai.mmlu_pro.business.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 84.664,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "business",
+          "task_name": "Business",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "84.664",
+          "raw_stderr": "1.779",
+          "latency": "11.164",
+          "cost_per_test": "0.001254",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.779,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:chemistry:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.chemistry",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Chemistry",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "chemistry",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Chemistry).",
+        "metric_id": "vals_ai.mmlu_pro.chemistry.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 84.982,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "chemistry",
+          "task_name": "Chemistry",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "84.982",
+          "raw_stderr": "1.375",
+          "latency": "15.68",
+          "cost_per_test": "0.001866",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.375,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:computer_science:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.computer_science",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Computer Science",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "computer_science",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Computer Science).",
+        "metric_id": "vals_ai.mmlu_pro.computer_science.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 83.171,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "computer_science",
+          "task_name": "Computer Science",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "83.171",
+          "raw_stderr": "2.247",
+          "latency": "11.872",
+          "cost_per_test": "0.00154",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.247,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:economics:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.economics",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Economics",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "economics",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Economics).",
+        "metric_id": "vals_ai.mmlu_pro.economics.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 85.782,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "economics",
+          "task_name": "Economics",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "85.782",
+          "raw_stderr": "1.662",
+          "latency": "11.545",
+          "cost_per_test": "0.001255",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.662,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:engineering:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.engineering",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Engineering",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "engineering",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Engineering).",
+        "metric_id": "vals_ai.mmlu_pro.engineering.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 74.923,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "engineering",
+          "task_name": "Engineering",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "74.923",
+          "raw_stderr": "1.595",
+          "latency": "26.843",
+          "cost_per_test": "0.0025",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.595,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:health:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.health",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Health",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "health",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Health).",
+        "metric_id": "vals_ai.mmlu_pro.health.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 77.262,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "health",
+          "task_name": "Health",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "77.262",
+          "raw_stderr": "1.738",
+          "latency": "11.318",
+          "cost_per_test": "0.00113",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.738,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:history:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.history",
+      "source_data": {
+        "dataset_name": "MMLU Pro - History",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "history",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (History).",
+        "metric_id": "vals_ai.mmlu_pro.history.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 70.866,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "history",
+          "task_name": "History",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "70.866",
+          "raw_stderr": "2.555",
+          "latency": "11.778",
+          "cost_per_test": "0.001668",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.555,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:law:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.law",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Law",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "law",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Law).",
+        "metric_id": "vals_ai.mmlu_pro.law.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 56.403,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "law",
+          "task_name": "Law",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "56.403",
+          "raw_stderr": "1.306",
+          "latency": "17.014",
+          "cost_per_test": "0.001607",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.306,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:math:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.math",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Math",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "math",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Math).",
+        "metric_id": "vals_ai.mmlu_pro.math.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 88.379,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "math",
+          "task_name": "Math",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "88.379",
+          "raw_stderr": "1.232",
+          "latency": "15.538",
+          "cost_per_test": "0.001913",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.232,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:other:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.other",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Others",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "other",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Others).",
+        "metric_id": "vals_ai.mmlu_pro.other.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 79.221,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "other",
+          "task_name": "Others",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "79.221",
+          "raw_stderr": "1.632",
+          "latency": "8.956",
+          "cost_per_test": "0.000929",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.632,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:overall:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.overall",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Overall",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "overall",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Overall).",
+        "metric_id": "vals_ai.mmlu_pro.overall.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 79.823,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "overall",
+          "task_name": "Overall",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "79.823",
+          "raw_stderr": "0.472",
+          "latency": "13.554",
+          "cost_per_test": "0.001483",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 0.472,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:philosophy:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.philosophy",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Philosophy",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "philosophy",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Philosophy).",
+        "metric_id": "vals_ai.mmlu_pro.philosophy.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 76.353,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "philosophy",
+          "task_name": "Philosophy",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "76.353",
+          "raw_stderr": "2.169",
+          "latency": "9.977",
+          "cost_per_test": "0.001026",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 2.169,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:physics:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.physics",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Physics",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "physics",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Physics).",
+        "metric_id": "vals_ai.mmlu_pro.physics.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 86.374,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "physics",
+          "task_name": "Physics",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "86.374",
+          "raw_stderr": "1.213",
+          "latency": "14.859",
+          "cost_per_test": "0.001607",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.213,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    },
+    {
+      "evaluation_result_id": "mmlu_pro:psychology:mistralai/mistral-large-2512:score",
+      "evaluation_name": "vals_ai.mmlu_pro.psychology",
+      "source_data": {
+        "dataset_name": "MMLU Pro - Psychology",
+        "source_type": "url",
+        "url": [
+          "https://www.vals.ai/benchmarks/mmlu_pro"
+        ],
+        "additional_details": {
+          "benchmark_slug": "mmlu_pro",
+          "task_key": "psychology",
+          "dataset_type": "public",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "metric_config": {
+        "evaluation_description": "Accuracy reported by Vals.ai for MMLU Pro (Psychology).",
+        "metric_id": "vals_ai.mmlu_pro.psychology.accuracy",
+        "metric_name": "Accuracy",
+        "metric_kind": "accuracy",
+        "metric_unit": "percent",
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "additional_details": {
+          "score_scale": "percent_0_to_100",
+          "max_score_source": "fixed_percentage_bound",
+          "leaderboard_page_url": "https://www.vals.ai/benchmarks/mmlu_pro"
+        }
+      },
+      "score_details": {
+        "score": 80.576,
+        "details": {
+          "benchmark_slug": "mmlu_pro",
+          "benchmark_name": "MMLU Pro",
+          "benchmark_updated": "2026-04-23",
+          "task_key": "psychology",
+          "task_name": "Psychology",
+          "dataset_type": "public",
+          "industry": "academic",
+          "raw_score": "80.576",
+          "raw_stderr": "1.754",
+          "latency": "10.746",
+          "cost_per_test": "0.001079",
+          "temperature": "0.065",
+          "max_output_tokens": "256000",
+          "provider": "Mistral AI"
+        },
+        "uncertainty": {
+          "standard_error": {
+            "value": 1.754,
+            "method": "vals_reported"
+          }
+        }
+      },
+      "generation_config": {
+        "generation_args": {
+          "temperature": 0.065,
+          "max_tokens": 256000,
+          "max_attempts": 1
+        }
+      }
+    }
+  ]
+}

flat/objects/00/59/0059b10e-20a1-4543-be26-bfd6cc82c33e.json

@@ -0,0 +1,178 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "MMLongBench/InternVL3-2B/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "University of Edinburgh",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "InternVL3-2B",
+    "name": "InternVL3-2B",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "MMLongBench",
+      "source_data": {
+        "dataset_name": "MMLongBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2505.10610"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall average performance score across all five task categories (VRAG, NIAH, ICL, Summarization, and DocVQA) on the MMLONGBENCH benchmark. This score represents a holistic measure of a model's long-context vision-language capabilities at the most challenging 128K token context length. Scores are percentages, with higher being better. Full results from Figure 23 in the paper.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Average Score (%) @ 128K",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Overall Average Performance on MMLONGBENCH (128K Tokens)"
+        },
+        "metric_id": "overall_average_performance_on_mmlongbench_128k_tokens",
+        "metric_name": "Overall Average Performance on MMLONGBENCH (128K Tokens)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 18.2
+      },
+      "evaluation_result_id": "MMLongBench/InternVL3-2B/1771591481.616601#mmlongbench#overall_average_performance_on_mmlongbench_128k_tokens"
+    },
+    {
+      "evaluation_name": "MMLongBench",
+      "source_data": {
+        "dataset_name": "MMLongBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2505.10610"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall average performance score across all five task categories (VRAG, NIAH, ICL, Summarization, and DocVQA) on the MMLONGBENCH benchmark. This score represents a holistic measure of a model's long-context vision-language capabilities at the 16K token context length. Scores are percentages, with higher being better. Full results from Figure 23 in the paper.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Average Score (%) @ 16K",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Average Performance on MMLONGBENCH (16K Tokens)"
+        },
+        "metric_id": "overall_average_performance_on_mmlongbench_16k_tokens",
+        "metric_name": "Overall Average Performance on MMLONGBENCH (16K Tokens)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 35.1
+      },
+      "evaluation_result_id": "MMLongBench/InternVL3-2B/1771591481.616601#mmlongbench#overall_average_performance_on_mmlongbench_16k_tokens"
+    },
+    {
+      "evaluation_name": "MMLongBench",
+      "source_data": {
+        "dataset_name": "MMLongBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2505.10610"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall average performance score across all five task categories (VRAG, NIAH, ICL, Summarization, and DocVQA) on the MMLONGBENCH benchmark. This score represents a holistic measure of a model's long-context vision-language capabilities at the 32K token context length. Scores are percentages, with higher being better. Full results from Figure 23 in the paper.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Average Score (%) @ 32K",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Average Performance on MMLONGBENCH (32K Tokens)"
+        },
+        "metric_id": "overall_average_performance_on_mmlongbench_32k_tokens",
+        "metric_name": "Overall Average Performance on MMLONGBENCH (32K Tokens)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 26.5
+      },
+      "evaluation_result_id": "MMLongBench/InternVL3-2B/1771591481.616601#mmlongbench#overall_average_performance_on_mmlongbench_32k_tokens"
+    },
+    {
+      "evaluation_name": "MMLongBench",
+      "source_data": {
+        "dataset_name": "MMLongBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2505.10610"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall average performance score across all five task categories (VRAG, NIAH, ICL, Summarization, and DocVQA) on the MMLONGBENCH benchmark. This score represents a holistic measure of a model's long-context vision-language capabilities at the 64K token context length. Scores are percentages, with higher being better. Full results from Figure 23 in the paper.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Average Score (%) @ 64K",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Average Performance on MMLONGBENCH (64K Tokens)"
+        },
+        "metric_id": "overall_average_performance_on_mmlongbench_64k_tokens",
+        "metric_name": "Overall Average Performance on MMLONGBENCH (64K Tokens)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 21.9
+      },
+      "evaluation_result_id": "MMLongBench/InternVL3-2B/1771591481.616601#mmlongbench#overall_average_performance_on_mmlongbench_64k_tokens"
+    },
+    {
+      "evaluation_name": "MMLongBench",
+      "source_data": {
+        "dataset_name": "MMLongBench",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2505.10610"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall average performance score across all five task categories (VRAG, NIAH, ICL, Summarization, and DocVQA) on the MMLONGBENCH benchmark. This score represents a holistic measure of a model's long-context vision-language capabilities at the 8K token context length. Scores are percentages, with higher being better. Full results from Figure 23 in the paper.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Average Score (%) @ 8K",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Average Performance on MMLONGBENCH (8K Tokens)"
+        },
+        "metric_id": "overall_average_performance_on_mmlongbench_8k_tokens",
+        "metric_name": "Overall Average Performance on MMLONGBENCH (8K Tokens)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 45.8
+      },
+      "evaluation_result_id": "MMLongBench/InternVL3-2B/1771591481.616601#mmlongbench#overall_average_performance_on_mmlongbench_8k_tokens"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/5b/005b6666-57ce-4fb6-a9d9-1b1ccc8e354d.json

@@ -0,0 +1,148 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "IFEvalCode/GPT-4o/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Beihang University",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "GPT-4o",
+    "name": "GPT-4o",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "IFEvalCode",
+      "source_data": {
+        "dataset_name": "IFEvalCode",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2507.22462"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average instruction-following Pass@1 score across 8 programming languages (Python, Java, C++, C#, TypeScript, JavaScript, PHP, Shell) on the IFEvalCode benchmark using English queries. This metric measures the percentage of generated code snippets that correctly adhere to all human-provided constraints (e.g., style, variable naming, algorithm usage), which is the key innovation of this benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Instruction-Following Avg. Pass@1 (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "IFEvalCode: Average Instruction-Following on English Queries"
+        },
+        "metric_id": "ifevalcode_average_instruction_following_on_english_queries",
+        "metric_name": "IFEvalCode: Average Instruction-Following on English Queries",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 20.5
+      },
+      "evaluation_result_id": "IFEvalCode/GPT-4o/1771591481.616601#ifevalcode#ifevalcode_average_instruction_following_on_english_queries"
+    },
+    {
+      "evaluation_name": "IFEvalCode",
+      "source_data": {
+        "dataset_name": "IFEvalCode",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2507.22462"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average correctness Pass@1 score across 8 programming languages (Python, Java, C++, C#, TypeScript, JavaScript, PHP, Shell) on the IFEvalCode benchmark using Chinese queries. This metric measures the percentage of generated code snippets that are functionally correct and pass standard unit tests, testing the multilingual capabilities of models.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Correctness Avg. Pass@1 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "IFEvalCode: Average Correctness on Chinese Queries"
+        },
+        "metric_id": "ifevalcode_average_correctness_on_chinese_queries",
+        "metric_name": "IFEvalCode: Average Correctness on Chinese Queries",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 32.2
+      },
+      "evaluation_result_id": "IFEvalCode/GPT-4o/1771591481.616601#ifevalcode#ifevalcode_average_correctness_on_chinese_queries"
+    },
+    {
+      "evaluation_name": "IFEvalCode",
+      "source_data": {
+        "dataset_name": "IFEvalCode",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2507.22462"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average correctness Pass@1 score across 8 programming languages (Python, Java, C++, C#, TypeScript, JavaScript, PHP, Shell) on the IFEvalCode benchmark using English queries. This metric measures the percentage of generated code snippets that are functionally correct and pass standard unit tests.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Correctness Avg. Pass@1 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "IFEvalCode: Average Correctness on English Queries"
+        },
+        "metric_id": "ifevalcode_average_correctness_on_english_queries",
+        "metric_name": "IFEvalCode: Average Correctness on English Queries",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 33.6
+      },
+      "evaluation_result_id": "IFEvalCode/GPT-4o/1771591481.616601#ifevalcode#ifevalcode_average_correctness_on_english_queries"
+    },
+    {
+      "evaluation_name": "IFEvalCode",
+      "source_data": {
+        "dataset_name": "IFEvalCode",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2507.22462"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average instruction-following Pass@1 score across 8 programming languages (Python, Java, C++, C#, TypeScript, JavaScript, PHP, Shell) on the IFEvalCode benchmark using Chinese queries. This metric measures the percentage of generated code snippets that correctly adhere to all human-provided constraints (e.g., style, variable naming, algorithm usage), testing the multilingual capabilities of models.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Instruction-Following Avg. Pass@1 (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "IFEvalCode: Average Instruction-Following on Chinese Queries"
+        },
+        "metric_id": "ifevalcode_average_instruction_following_on_chinese_queries",
+        "metric_name": "IFEvalCode: Average Instruction-Following on Chinese Queries",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 22
+      },
+      "evaluation_result_id": "IFEvalCode/GPT-4o/1771591481.616601#ifevalcode#ifevalcode_average_instruction_following_on_chinese_queries"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/5c/005cc577-3c03-40e8-bcb8-4c34abcc31e1.json

@@ -0,0 +1,58 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "VisDrone2018/Transformer/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Tianjin University",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Transformer",
+    "name": "Transformer",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "VisDrone2018",
+      "source_data": {
+        "dataset_name": "VisDrone2018",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/1804.07437"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": true,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "This benchmark evaluates crowd counting models on the VisDrone2018 dataset, based on the 2022 challenge leaderboard. The metric is Mean Absolute Error (MAE), which measures the average absolute difference between the predicted and actual number of people in an image. Lower MAE scores indicate better performance.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MAE",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Crowd Counting on VisDrone2018 (2022 Leaderboard)"
+        },
+        "metric_id": "crowd_counting_on_visdrone2018_2022_leaderboard",
+        "metric_name": "Crowd Counting on VisDrone2018 (2022 Leaderboard)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 15.95
+      },
+      "evaluation_result_id": "VisDrone2018/Transformer/1771591481.616601#visdrone2018#crowd_counting_on_visdrone2018_2022_leaderboard"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/5e/005e54ae-9901-448d-834e-fe4e6d134b50.json

@@ -0,0 +1,58 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "MMMU-Pro/Skywork-R1V2-38B/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Carnegie Mellon University",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Skywork-R1V2-38B",
+    "name": "Skywork-R1V2-38B",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "MMMU-Pro",
+      "source_data": {
+        "dataset_name": "MMMU-Pro",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2409.02813"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Overall accuracy on the original MMMU validation set. This serves as a baseline to compare against the more robust MMMU-Pro benchmark and shows the performance drop when shortcuts are mitigated.",
+        "additional_details": {
+          "alphaxiv_y_axis": "MMMU(Val) Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "MMMU Benchmark (Validation Set)"
+        },
+        "metric_id": "mmmu_benchmark_validation_set",
+        "metric_name": "MMMU Benchmark (Validation Set)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 73.6
+      },
+      "evaluation_result_id": "MMMU-Pro/Skywork-R1V2-38B/1771591481.616601#mmmu_pro#mmmu_benchmark_validation_set"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/5e/005eaac6-f703-454d-8716-5bc998a329b9.json

@@ -0,0 +1,238 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "CRIC/Memory-VQA+l_att/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Chinese Academy of Sciences",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "Memory-VQA+l_att",
+    "name": "Memory-VQA+l_att",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "CRIC",
+      "source_data": {
+        "dataset_name": "CRIC",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/1908.02962"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of questions on the CRIC test set where both the answer and the object grounding are correct. This is the primary metric for the CRIC benchmark, as it evaluates a model's holistic compositional reasoning ability by requiring success in both sub-tasks.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Final Score (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Overall Final Score on the CRIC Benchmark"
+        },
+        "metric_id": "overall_final_score_on_the_cric_benchmark",
+        "metric_name": "Overall Final Score on the CRIC Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 38.87
+      },
+      "evaluation_result_id": "CRIC/Memory-VQA+l_att/1771591481.616601#cric#overall_final_score_on_the_cric_benchmark"
+    },
+    {
+      "evaluation_name": "CRIC",
+      "source_data": {
+        "dataset_name": "CRIC",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/1908.02962"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the overall percentage of correctly grounded target objects across all question types on the CRIC VQA benchmark. This metric highlights a key challenge of the dataset.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Grounding Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Grounding Accuracy on the CRIC Benchmark"
+        },
+        "metric_id": "overall_grounding_accuracy_on_the_cric_benchmark",
+        "metric_name": "Overall Grounding Accuracy on the CRIC Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 49.51
+      },
+      "evaluation_result_id": "CRIC/Memory-VQA+l_att/1771591481.616601#cric#overall_grounding_accuracy_on_the_cric_benchmark"
+    },
+    {
+      "evaluation_name": "CRIC",
+      "source_data": {
+        "dataset_name": "CRIC",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/1908.02962"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the overall percentage of correctly answered questions across all types on the CRIC VQA benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Overall Answer Accuracy (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Answer Accuracy on the CRIC Benchmark"
+        },
+        "metric_id": "overall_answer_accuracy_on_the_cric_benchmark",
+        "metric_name": "Overall Answer Accuracy on the CRIC Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 66.93
+      },
+      "evaluation_result_id": "CRIC/Memory-VQA+l_att/1771591481.616601#cric#overall_answer_accuracy_on_the_cric_benchmark"
+    },
+    {
+      "evaluation_name": "CRIC",
+      "source_data": {
+        "dataset_name": "CRIC",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/1908.02962"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of correctly grounded target objects for 'Recognize' (non-yes/no) questions on the CRIC VQA benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Grounding Accuracy (%) - Recognize Questions",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Grounding Accuracy on 'Recognize' Questions in CRIC"
+        },
+        "metric_id": "grounding_accuracy_on_recognize_questions_in_cric",
+        "metric_name": "Grounding Accuracy on 'Recognize' Questions in CRIC",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 44.65
+      },
+      "evaluation_result_id": "CRIC/Memory-VQA+l_att/1771591481.616601#cric#grounding_accuracy_on_recognize_questions_in_cric"
+    },
+    {
+      "evaluation_name": "CRIC",
+      "source_data": {
+        "dataset_name": "CRIC",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/1908.02962"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of correctly answered 'Verify' (yes/no) questions on the CRIC VQA benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Answer Accuracy (%) - Verify Questions",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Answer Accuracy on 'Verify' Questions in CRIC"
+        },
+        "metric_id": "answer_accuracy_on_verify_questions_in_cric",
+        "metric_name": "Answer Accuracy on 'Verify' Questions in CRIC",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 77.44
+      },
+      "evaluation_result_id": "CRIC/Memory-VQA+l_att/1771591481.616601#cric#answer_accuracy_on_verify_questions_in_cric"
+    },
+    {
+      "evaluation_name": "CRIC",
+      "source_data": {
+        "dataset_name": "CRIC",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/1908.02962"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of correctly grounded target objects for 'Verify' (yes/no) questions on the CRIC VQA benchmark. For 'no' answers, grounding is considered correct if the model correctly indicates 'no object'.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Grounding Accuracy (%) - Verify Questions",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Grounding Accuracy on 'Verify' Questions in CRIC"
+        },
+        "metric_id": "grounding_accuracy_on_verify_questions_in_cric",
+        "metric_name": "Grounding Accuracy on 'Verify' Questions in CRIC",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 61.39
+      },
+      "evaluation_result_id": "CRIC/Memory-VQA+l_att/1771591481.616601#cric#grounding_accuracy_on_verify_questions_in_cric"
+    },
+    {
+      "evaluation_name": "CRIC",
+      "source_data": {
+        "dataset_name": "CRIC",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/1908.02962"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Measures the percentage of correctly answered 'Recognize' (non-yes/no) questions on the CRIC VQA benchmark.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Answer Accuracy (%) - Recognize Questions",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Answer Accuracy on 'Recognize' Questions in CRIC"
+        },
+        "metric_id": "answer_accuracy_on_recognize_questions_in_cric",
+        "metric_name": "Answer Accuracy on 'Recognize' Questions in CRIC",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 62.64
+      },
+      "evaluation_result_id": "CRIC/Memory-VQA+l_att/1771591481.616601#cric#answer_accuracy_on_recognize_questions_in_cric"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/60/006051e1-9290-4492-a5f8-1ae16e0db19f.json

@@ -0,0 +1,478 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "OCRBench v2/GPT-4o/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "South China University of Technology",
+      "alphaxiv_dataset_type": "image",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "GPT-4o",
+    "name": "GPT-4o",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average score across eight core OCR capabilities (Recognition, Referring, Spotting, Extraction, Parsing, Calculation, Understanding, and Reasoning) on the private English test set of OCRBench v2. This metric provides a holistic view of a model's ability to handle diverse text-centric visual tasks.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Average Score (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Overall Performance on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "overall_performance_on_ocrbench_v2_english_private",
+        "metric_name": "Overall Performance on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 47.6
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#overall_performance_on_ocrbench_v2_english_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Relation Extraction capability, which involves extracting and mapping visual components from densely arranged text, such as key-value pairs from receipts or handwritten content. It is evaluated using an F1 score on the private Chinese test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Relation Extraction Score (Chinese, %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Relation Extraction on OCRBench v2 (Chinese, Private)"
+        },
+        "metric_id": "relation_extraction_on_ocrbench_v2_chinese_private",
+        "metric_name": "Relation Extraction on OCRBench v2 (Chinese, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 52.1
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#relation_extraction_on_ocrbench_v2_chinese_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Element Parsing capability, assessing a model's ability to parse complex elements like tables, charts, and documents into structured formats (e.g., HTML, JSON). It is evaluated using metrics like Tree-Edit-Distance-based Similarity (TEDS) on the private Chinese test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Element Parsing Score (Chinese, %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Element Parsing on OCRBench v2 (Chinese, Private)"
+        },
+        "metric_id": "element_parsing_on_ocrbench_v2_chinese_private",
+        "metric_name": "Element Parsing on OCRBench v2 (Chinese, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 29
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#element_parsing_on_ocrbench_v2_chinese_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Knowledge Reasoning capability, which requires complex inference and world knowledge. Tasks include Science QA, APP Agent interaction, ASCII art classification, and Reasoning VQA, evaluated on the private Chinese test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Knowledge Reasoning Score (Chinese, %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Knowledge Reasoning on OCRBench v2 (Chinese, Private)"
+        },
+        "metric_id": "knowledge_reasoning_on_ocrbench_v2_chinese_private",
+        "metric_name": "Knowledge Reasoning on OCRBench v2 (Chinese, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 29.4
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#knowledge_reasoning_on_ocrbench_v2_chinese_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Text Recognition capability, which assesses a model's ability to perceive and transcribe textual content from images. This includes tasks like recognizing regular, irregular, handwritten, and occluded text on the private Chinese test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Text Recognition Score (Chinese, %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Text Recognition on OCRBench v2 (Chinese, Private)"
+        },
+        "metric_id": "text_recognition_on_ocrbench_v2_chinese_private",
+        "metric_name": "Text Recognition on OCRBench v2 (Chinese, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 41.7
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#text_recognition_on_ocrbench_v2_chinese_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Visual Text Understanding capability, which assesses a model's ability to comprehend the semantic information of text in an image. Tasks include document classification, diagram question answering, and cognition VQA, evaluated on the private Chinese test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Visual Text Understanding Score (Chinese, %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Visual Text Understanding on OCRBench v2 (Chinese, Private)"
+        },
+        "metric_id": "visual_text_understanding_on_ocrbench_v2_chinese_private",
+        "metric_name": "Visual Text Understanding on OCRBench v2 (Chinese, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 76
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#visual_text_understanding_on_ocrbench_v2_chinese_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Mathematical Calculation capability, which covers numerical reasoning tasks such as text counting (character/word counts) and solving math problems presented in images (Math QA). Scores are based on exact match or normalized L1 distance on the private English test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Mathematical Calculation Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Mathematical Calculation on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "mathematical_calculation_on_ocrbench_v2_english_private",
+        "metric_name": "Mathematical Calculation on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 51.6
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#mathematical_calculation_on_ocrbench_v2_english_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Relation Extraction capability, which involves extracting and mapping visual components from densely arranged text, such as key-value pairs from receipts or handwritten content. It is evaluated using an F1 score on the private English test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Relation Extraction Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Relation Extraction on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "relation_extraction_on_ocrbench_v2_english_private",
+        "metric_name": "Relation Extraction on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 87.4
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#relation_extraction_on_ocrbench_v2_english_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Element Parsing capability, assessing a model's ability to parse complex elements like tables, charts, and documents into structured formats (e.g., HTML, JSON). It is evaluated using metrics like Tree-Edit-Distance-based Similarity (TEDS) on the private English test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Element Parsing Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Element Parsing on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "element_parsing_on_ocrbench_v2_english_private",
+        "metric_name": "Element Parsing on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 23.1
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#element_parsing_on_ocrbench_v2_english_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Knowledge Reasoning capability, which requires complex inference and world knowledge. Tasks include Science QA, APP Agent interaction, ASCII art classification, and Reasoning VQA, evaluated on the private English test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Knowledge Reasoning Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Knowledge Reasoning on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "knowledge_reasoning_on_ocrbench_v2_english_private",
+        "metric_name": "Knowledge Reasoning on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 62.3
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#knowledge_reasoning_on_ocrbench_v2_english_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Text Recognition capability, which assesses a model's ability to perceive and transcribe textual content from images. This includes tasks like recognizing regular, irregular, handwritten, and occluded text on the private English test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Text Recognition Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Text Recognition on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "text_recognition_on_ocrbench_v2_english_private",
+        "metric_name": "Text Recognition on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 58.6
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#text_recognition_on_ocrbench_v2_english_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Text Referring capability, which measures a model's ability to accurately determine the location of specified text within an image. This often involves outputting bounding box coordinates and is evaluated using metrics like Intersection over Union (IoU) on the private English test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Text Referring Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Text Referring on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "text_referring_on_ocrbench_v2_english_private",
+        "metric_name": "Text Referring on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 23.4
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#text_referring_on_ocrbench_v2_english_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Text Spotting capability, which requires models to both locate (output bounding box) and recognize (transcribe content) text simultaneously. It is a challenging task where most LMMs perform poorly, evaluated using an F1 score on the private English test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Text Spotting Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Text Spotting on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "text_spotting_on_ocrbench_v2_english_private",
+        "metric_name": "Text Spotting on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 0
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#text_spotting_on_ocrbench_v2_english_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average score across five core OCR capabilities (Recognition, Extraction, Parsing, Understanding, and Reasoning) on the private Chinese test set of OCRBench v2. This metric provides a holistic view of a model's ability to handle diverse Chinese text-centric visual tasks.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Average Score (Chinese, %)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Overall Performance on OCRBench v2 (Chinese, Private)"
+        },
+        "metric_id": "overall_performance_on_ocrbench_v2_chinese_private",
+        "metric_name": "Overall Performance on OCRBench v2 (Chinese, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 45.7
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#overall_performance_on_ocrbench_v2_chinese_private"
+    },
+    {
+      "evaluation_name": "OCRBench v2",
+      "source_data": {
+        "dataset_name": "OCRBench v2",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2501.00321"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Performance on the Visual Text Understanding capability, which assesses a model's ability to comprehend the semantic information of text in an image. Tasks include document classification, diagram question answering, and cognition VQA, evaluated on the private English test set.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Visual Text Understanding Score (%)",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Visual Text Understanding on OCRBench v2 (English, Private)"
+        },
+        "metric_id": "visual_text_understanding_on_ocrbench_v2_english_private",
+        "metric_name": "Visual Text Understanding on OCRBench v2 (English, Private)",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 74.4
+      },
+      "evaluation_result_id": "OCRBench v2/GPT-4o/1771591481.616601#ocrbench_v2#visual_text_understanding_on_ocrbench_v2_english_private"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}

flat/objects/00/61/0061098c-b43d-47fd-b6ca-c027a8e70b8e.json

@@ -0,0 +1,178 @@
+{
+  "schema_version": "0.2.2",
+  "evaluation_id": "MCLM/GPT-4o-Mini/1771591481.616601",
+  "retrieved_timestamp": "1771591481.616601",
+  "source_metadata": {
+    "source_name": "alphaXiv State of the Art",
+    "source_type": "documentation",
+    "source_organization_name": "alphaXiv",
+    "source_organization_url": "https://alphaxiv.org",
+    "evaluator_relationship": "third_party",
+    "additional_details": {
+      "alphaxiv_dataset_org": "Yonsei University",
+      "alphaxiv_dataset_type": "text",
+      "scrape_source": "https://github.com/alphaXiv/feedback/issues/189"
+    }
+  },
+  "model_info": {
+    "id": "GPT-4o-Mini",
+    "name": "GPT-4o-Mini",
+    "developer": "unknown"
+  },
+  "evaluation_results": [
+    {
+      "evaluation_name": "MCLM",
+      "source_data": {
+        "dataset_name": "MCLM",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.17407"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Average accuracy across all four subsets (MT-MATH100, MT-AIME2024, M-IMO, M-MO) of the Multilingual Competition Level Math (MCLM) benchmark. This metric, from Table 4, represents the overall mathematical reasoning capability of models in a challenging, multilingual setting.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Average Accuracy (%)",
+          "alphaxiv_is_primary": "True",
+          "raw_evaluation_name": "Overall Performance on the MCLM Benchmark"
+        },
+        "metric_id": "overall_performance_on_the_mclm_benchmark",
+        "metric_name": "Overall Performance on the MCLM Benchmark",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 33.66
+      },
+      "evaluation_result_id": "MCLM/GPT-4o-Mini/1771591481.616601#mclm#overall_performance_on_the_mclm_benchmark"
+    },
+    {
+      "evaluation_name": "MCLM",
+      "source_data": {
+        "dataset_name": "MCLM",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.17407"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the M-IMO subset of the MCLM benchmark, evaluated by an LLM-as-a-Judge (from Table 4). This subset consists of problems from the International Mathematical Olympiad (2006, 2024), covering 38 languages.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) on M-IMO",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Performance on MCLM's M-IMO Subset"
+        },
+        "metric_id": "performance_on_mclm_s_m_imo_subset",
+        "metric_name": "Performance on MCLM's M-IMO Subset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 13.33
+      },
+      "evaluation_result_id": "MCLM/GPT-4o-Mini/1771591481.616601#mclm#performance_on_mclm_s_m_imo_subset"
+    },
+    {
+      "evaluation_name": "MCLM",
+      "source_data": {
+        "dataset_name": "MCLM",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.17407"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the MT-AIME2024 subset of the MCLM benchmark, from Table 4. This subset consists of 30 challenging problems from the AIME 2024 competition, machine-translated into 55 languages.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) on MT-AIME2024",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Performance on MCLM's MT-AIME2024 Subset"
+        },
+        "metric_id": "performance_on_mclm_s_mt_aime2024_subset",
+        "metric_name": "Performance on MCLM's MT-AIME2024 Subset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 20.18
+      },
+      "evaluation_result_id": "MCLM/GPT-4o-Mini/1771591481.616601#mclm#performance_on_mclm_s_mt_aime2024_subset"
+    },
+    {
+      "evaluation_name": "MCLM",
+      "source_data": {
+        "dataset_name": "MCLM",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.17407"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the MT-MATH100 subset of the MCLM benchmark, from Table 4. This subset consists of 100 competition-level problems from the Math-500 dataset, machine-translated into 55 languages.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) on MT-MATH100",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Performance on MCLM's MT-MATH100 Subset"
+        },
+        "metric_id": "performance_on_mclm_s_mt_math100_subset",
+        "metric_name": "Performance on MCLM's MT-MATH100 Subset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 70.3
+      },
+      "evaluation_result_id": "MCLM/GPT-4o-Mini/1771591481.616601#mclm#performance_on_mclm_s_mt_math100_subset"
+    },
+    {
+      "evaluation_name": "MCLM",
+      "source_data": {
+        "dataset_name": "MCLM",
+        "source_type": "url",
+        "url": [
+          "https://www.alphaxiv.org/abs/2502.17407"
+        ]
+      },
+      "metric_config": {
+        "lower_is_better": false,
+        "score_type": "continuous",
+        "min_score": 0.0,
+        "max_score": 100.0,
+        "evaluation_description": "Accuracy on the M-MO subset of the MCLM benchmark, evaluated by an LLM-as-a-Judge (from Table 4). This subset features problems from various domestic and regional math Olympiads across 11 languages.",
+        "additional_details": {
+          "alphaxiv_y_axis": "Accuracy (%) on M-MO",
+          "alphaxiv_is_primary": "False",
+          "raw_evaluation_name": "Performance on MCLM's M-MO Subset"
+        },
+        "metric_id": "performance_on_mclm_s_m_mo_subset",
+        "metric_name": "Performance on MCLM's M-MO Subset",
+        "metric_kind": "score",
+        "metric_unit": "points"
+      },
+      "score_details": {
+        "score": 30.81
+      },
+      "evaluation_result_id": "MCLM/GPT-4o-Mini/1771591481.616601#mclm#performance_on_mclm_s_m_mo_subset"
+    }
+  ],
+  "eval_library": {
+    "name": "alphaxiv",
+    "version": "unknown"
+  }
+}