data/benchmarks/categories.json

{
  "AA-Index": [
    "knowledge",
    "general"
  ],
  "AA-LCR": [
    "general",
    "applied_reasoning"
  ],
  "ACE": [
    "multimodal",
    "agentic"
  ],
  "ACEBench": [
    "multimodal",
    "agentic"
  ],
  "AGIEval": [
    "general",
    "applied_reasoning"
  ],
  "AI2 Reasoning Challenge (ARC)": [
    "logical_reasoning",
    "knowledge",
    "commonsense_reasoning"
  ],
  "AI2D": [
    "multimodal",
    "knowledge"
  ],
  "AIME": [
    "mathematics"
  ],
  "AIME 2024": [
    "mathematics"
  ],
  "AIME 2025": [
    "mathematics"
  ],
  "AITZ_EM": [
    "agentic"
  ],
  "AMC_2022_23": [
    "mathematics"
  ],
  "APEX Agents": [
    "agentic",
    "applied_reasoning"
  ],
  "APEX v1": [
    "agentic",
    "applied_reasoning"
  ],
  "API-Bank": [
    "agentic"
  ],
  "Air Bench 2024": [
    "safety"
  ],
  "ARC-AGI": [
    "logical_reasoning",
    "commonsense_reasoning",
    "general"
  ],
  "ARC-C": [
    "logical_reasoning",
    "knowledge",
    "commonsense_reasoning"
  ],
  "ARC-E": [
    "logical_reasoning",
    "knowledge",
    "commonsense_reasoning"
  ],
  "ARKitScenes": [
    "multimodal"
  ],
  "ActivityNet": [
    "multimodal"
  ],
  "Agentharm": [
    "safety",
    "agentic"
  ],
  "Aider": [
    "software_engineering",
    "agentic"
  ],
  "Aider-Polyglot": [
    "software_engineering",
    "linguistic_core"
  ],
  "Aider-Polyglot Edit": [
    "software_engineering",
    "linguistic_core"
  ],
  "AlignBench": [
    "applied_reasoning",
    "general"
  ],
  "AlpacaEval 2.0": [
    "applied_reasoning",
    "general"
  ],
  "Android Control High_EM": [
    "agentic",
    "multimodal"
  ],
  "Android Control Low_EM": [
    "agentic",
    "multimodal"
  ],
  "AndroidWorld": [
    "agentic",
    "multimodal"
  ],
  "AndroidWorld_SR": [
    "agentic",
    "multimodal"
  ],
  "Anthropic Red Team": [
    "safety"
  ],
  "AppWorld": [
    "agentic"
  ],
  "AppWorld Benchmark": [
    "agentic",
    "applied_reasoning"
  ],
  "Arc": [
    "logical_reasoning",
    "commonsense_reasoning"
  ],
  "Arena Hard": [
    "general",
    "applied_reasoning"
  ],
  "Arena-Hard v2": [
    "general",
    "applied_reasoning"
  ],
  "Artificial Analysis": [
    "general"
  ],
  "Artificial Analysis LLM API": [
    "general"
  ],
  "Artificial-Analysis": [
    "general"
  ],
  "AssistantBench": [
    "agentic",
    "applied_reasoning"
  ],
  "AttaQ": [
    "robustness"
  ],
  "AutoLogi": [
    "logical_reasoning",
    "mathematics"
  ],
  "BBH": [
    "logical_reasoning",
    "applied_reasoning"
  ],
  "BBQ": [
    "safety",
    "commonsense_reasoning"
  ],
  "BFCL": [
    "agentic",
    "software_engineering"
  ],
  "BFCL v2": [
    "agentic",
    "software_engineering"
  ],
  "BFCL-V4": [
    "agentic",
    "software_engineering"
  ],
  "BFCL-v3": [
    "agentic",
    "software_engineering"
  ],
  "BFCL_v3_MultiTurn": [
    "agentic",
    "software_engineering"
  ],
  "BIG-Bench": [
    "general",
    "applied_reasoning"
  ],
  "BIG-Bench Extra Hard": [
    "general",
    "applied_reasoning"
  ],
  "BIG-Bench Hard": [
    "general",
    "applied_reasoning"
  ],
  "BLINK": [
    "knowledge",
    "linguistic_core"
  ],
  "BabyVision": [
    "multimodal"
  ],
  "Beyond AIME": [
    "mathematics"
  ],
  "BigCodeBench": [
    "software_engineering"
  ],
  "BigCodeBench-Full": [
    "software_engineering"
  ],
  "BigCodeBench-Hard": [
    "software_engineering"
  ],
  "BioLP-Bench": [
    "natural_sciences"
  ],
  "Bird-SQL (dev)": [
    "software_engineering"
  ],
  "BixBench": [
    "applied_reasoning"
  ],
  "BoolQ": [
    "commonsense_reasoning"
  ],
  "BrowseComp": [
    "agentic"
  ],
  "BrowseComp Long Context 128k": [
    "agentic"
  ],
  "BrowseComp Long Context 256k": [
    "agentic"
  ],
  "BrowseComp-Plus": [
    "agentic"
  ],
  "BrowseComp-VL": [
    "agentic",
    "multimodal"
  ],
  "BrowseComp-zh": [
    "agentic"
  ],
  "BrowseCompPlus": [
    "agentic"
  ],
  "C-Eval": [
    "knowledge",
    "mathematics",
    "natural_sciences",
    "humanities_and_social_sciences"
  ],
  "CBNSL": [
    "linguistic_core"
  ],
  "CC-Bench-V2 Backend": [
    "software_engineering"
  ],
  "CC-Bench-V2 Frontend": [
    "software_engineering"
  ],
  "CC-Bench-V2 Repo Exploration": [
    "software_engineering"
  ],
  "CC-OCR": [
    "multimodal"
  ],
  "CFEval": [
    "software_engineering"
  ],
  "CLUEWSC": [
    "linguistic_core"
  ],
  "CMMLU": [
    "knowledge",
    "mathematics",
    "natural_sciences",
    "humanities_and_social_sciences"
  ],
  "CNMO 2024": [
    "mathematics"
  ],
  "CNN/DailyMail": [
    "linguistic_core"
  ],
  "COLLIE": [
    "applied_reasoning"
  ],
  "CORE-Bench Hard": [
    "applied_reasoning"
  ],
  "CRAG": [
    "applied_reasoning"
  ],
  "CRPErelation": [
    "linguistic_core"
  ],
  "CRUX-O": [
    "software_engineering"
  ],
  "CRUXEval-Input-CoT": [
    "software_engineering"
  ],
  "CRUXEval-Output-CoT": [
    "software_engineering"
  ],
  "CSimpleQA": [
    "knowledge"
  ],
  "Caparena": [
    "agentic"
  ],
  "Caparena Auto": [
    "agentic"
  ],
  "CaseLaw (v2) - Overall": [
    "law"
  ],
  "CharXiv-D": [
    "multimodal"
  ],
  "CharXiv-R": [
    "multimodal"
  ],
  "CharadesSTA": [
    "multimodal"
  ],
  "ChartQA": [
    "multimodal"
  ],
  "CheXpert CXR": [
    "multimodal",
    "natural_sciences"
  ],
  "CivilComments": [
    "safety"
  ],
  "Claw-Eval": [
    "law"
  ],
  "CloningScenarios": [
    "robustness"
  ],
  "CoVoST2": [
    "linguistic_core"
  ],
  "CoVoST2 en-zh": [
    "linguistic_core"
  ],
  "CocoaBench": [
    "agentic"
  ],
  "CocoaBench v1.0": [
    "agentic"
  ],
  "CodeForces": [
    "mathematics",
    "software_engineering"
  ],
  "Codegolf v2.2": [
    "software_engineering"
  ],
  "Common Voice 15": [
    "linguistic_core"
  ],
  "CommonSenseQA": [
    "commonsense_reasoning"
  ],
  "Commonsense Qa": [
    "commonsense_reasoning"
  ],
  "ComplexFuncBench": [
    "software_engineering"
  ],
  "Corp Fin v2": [
    "finance"
  ],
  "CorpusQA 1M": [
    "knowledge"
  ],
  "CountBench": [
    "mathematics"
  ],
  "Creative Writing v3": [
    "linguistic_core"
  ],
  "CruxEval-O": [
    "software_engineering"
  ],
  "Cvebench": [
    "software_engineering",
    "safety"
  ],
  "CyBench": [
    "software_engineering",
    "safety"
  ],
  "Cybench": [
    "software_engineering",
    "safety"
  ],
  "CyberGym": [
    "agentic",
    "safety"
  ],
  "Cybersecurity CTFs": [
    "agentic",
    "safety"
  ],
  "Cyse2": [
    "software_engineering",
    "safety"
  ],
  "DROP": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "DS-Arena-Code": [
    "software_engineering"
  ],
  "DS-FIM-Eval": [
    "software_engineering"
  ],
  "DeepPlanning": [
    "agentic"
  ],
  "DeepSearchQA": [
    "applied_reasoning"
  ],
  "DermMCQA": [
    "natural_sciences"
  ],
  "Design2Code": [
    "software_engineering"
  ],
  "DocVQA": [
    "multimodal"
  ],
  "DocVQAtest": [
    "multimodal"
  ],
  "DynaMath": [
    "mathematics"
  ],
  "ECLeKTic": [
    "knowledge"
  ],
  "EQ-Bench": [
    "commonsense_reasoning"
  ],
  "ERQA": [
    "knowledge"
  ],
  "EgoSchema": [
    "multimodal"
  ],
  "EmbSpatialBench": [
    "commonsense_reasoning"
  ],
  "EvalPlus": [
    "software_engineering"
  ],
  "FActScore": [
    "hallucination"
  ],
  "FLEURS": [
    "linguistic_core"
  ],
  "FRAMES": [
    "linguistic_core"
  ],
  "Facts Grounding": [
    "hallucination"
  ],
  "Fibble Arena": [
    "logical_reasoning"
  ],
  "Fibble arena": [
    "logical_reasoning"
  ],
  "FigQA": [
    "multimodal"
  ],
  "FinQA": [
    "finance"
  ],
  "FinSearchComp T2&T3": [
    "finance"
  ],
  "FinSearchComp-T3": [
    "finance"
  ],
  "Finance Agent": [
    "agentic"
  ],
  "Flame-VLM-Code": [
    "software_engineering"
  ],
  "FlenQA": [
    "knowledge"
  ],
  "French MMLU": [
    "general"
  ],
  "FrontierMath": [
    "mathematics"
  ],
  "FrontierScience Research": [
    "natural_sciences"
  ],
  "FullStackBench en": [
    "software_engineering"
  ],
  "FullStackBench zh": [
    "software_engineering"
  ],
  "FunctionalMATH": [
    "mathematics"
  ],
  "GAIA": [
    "applied_reasoning",
    "agentic"
  ],
  "GDPval-AA": [
    "general"
  ],
  "GDPval-MM": [
    "multimodal"
  ],
  "GPQA": [
    "natural_sciences"
  ],
  "GPQA - Overall": [
    "natural_sciences"
  ],
  "GPQA Biology": [
    "natural_sciences"
  ],
  "GPQA Chemistry": [
    "natural_sciences"
  ],
  "GPQA Diamond": [
    "natural_sciences"
  ],
  "GPQA Physics": [
    "natural_sciences"
  ],
  "GSM-8K (CoT)": [
    "mathematics"
  ],
  "GSM-MC": [
    "mathematics"
  ],
  "GSM8K": [
    "mathematics"
  ],
  "GSM8K Chat": [
    "mathematics"
  ],
  "Gdm Intercode CTF": [
    "software_engineering",
    "agentic",
    "safety"
  ],
  "GeneBench": [
    "natural_sciences"
  ],
  "GiantSteps Tempo": [
    "multimodal"
  ],
  "Global MMLU Lite": [
    "general"
  ],
  "Global PIQA": [
    "commonsense_reasoning"
  ],
  "Global-MMLU": [
    "general"
  ],
  "Gorilla Benchmark API Bench": [
    "software_engineering",
    "agentic"
  ],
  "GovReport": [
    "linguistic_core"
  ],
  "Graphwalks BFS >128k": [
    "logical_reasoning"
  ],
  "Graphwalks parents >128k": [
    "logical_reasoning"
  ],
  "GroundUI-1K": [
    "multimodal",
    "agentic"
  ],
  "HAL": [
    "hallucination"
  ],
  "HELM": [
    "general"
  ],
  "HELM Instruct": [
    "general"
  ],
  "HF Open LLM Leaderboard v2": [
    "general"
  ],
  "HMMT 2025": [
    "mathematics"
  ],
  "HMMT Feb 26": [
    "mathematics"
  ],
  "HMMT25": [
    "mathematics"
  ],
  "Hallusion Bench": [
    "hallucination"
  ],
  "HarmBench": [
    "safety"
  ],
  "HealthBench": [
    "natural_sciences",
    "knowledge"
  ],
  "HealthBench Hard": [
    "natural_sciences",
    "knowledge"
  ],
  "HellaSwag": [
    "commonsense_reasoning"
  ],
  "Helm air bench": [
    "general"
  ],
  "Helm classic": [
    "general"
  ],
  "Helm lite": [
    "general"
  ],
  "HiddenMath": [
    "mathematics"
  ],
  "Holistic Evaluation of Language Models (HELM)": [
    "general"
  ],
  "HumanEval": [
    "software_engineering"
  ],
  "HumanEval-Average": [
    "software_engineering"
  ],
  "HumanEval-ER": [
    "software_engineering"
  ],
  "HumanEval-Mul": [
    "software_engineering"
  ],
  "HumanEvalFIM-Average": [
    "software_engineering"
  ],
  "Humanity's Last Exam": [
    "general"
  ],
  "Hypersim": [
    "multimodal"
  ],
  "IF": [
    "linguistic_core"
  ],
  "IFBench": [
    "linguistic_core"
  ],
  "IFEval": [
    "linguistic_core"
  ],
  "IMDb": [
    "linguistic_core"
  ],
  "IMO-AnswerBench": [
    "mathematics"
  ],
  "IOI": [
    "software_engineering"
  ],
  "IPhO 2025": [
    "natural_sciences",
    "mathematics"
  ],
  "ImageMining": [
    "multimodal"
  ],
  "Include": [
    "knowledge",
    "linguistic_core"
  ],
  "InfiniteBench/En.MC": [
    "linguistic_core"
  ],
  "InfiniteBench/En.QA": [
    "linguistic_core",
    "knowledge"
  ],
  "InfoVQA": [
    "multimodal",
    "knowledge"
  ],
  "InfoVQAtest": [
    "multimodal",
    "knowledge"
  ],
  "InfographicsQA": [
    "multimodal",
    "knowledge"
  ],
  "Instruct HumanEval": [
    "software_engineering"
  ],
  "InterGPS": [
    "multimodal"
  ],
  "Internal API instruction following (hard)": [
    "linguistic_core"
  ],
  "JudgeBench": [
    "applied_reasoning",
    "general"
  ],
  "Judgebench": [
    "applied_reasoning",
    "general"
  ],
  "LA Leaderboard": [
    "general"
  ],
  "LBPP (v2)": [
    "software_engineering"
  ],
  "LLM-Stats": [
    "general"
  ],
  "LMArena Text Leaderboard": [
    "general"
  ],
  "LSAT": [
    "law",
    "logical_reasoning"
  ],
  "LVBench": [
    "multimodal"
  ],
  "La Leaderboard composite dataset": [
    "general"
  ],
  "LegalBench": [
    "law"
  ],
  "LingoQA": [
    "linguistic_core",
    "knowledge"
  ],
  "Live-Bench": [
    "general"
  ],
  "LiveBench": [
    "general"
  ],
  "LiveBench 20241125": [
    "general"
  ],
  "LiveCodeBench": [
    "software_engineering"
  ],
  "LiveCodeBench - Easy": [
    "software_engineering"
  ],
  "LiveCodeBench - Hard": [
    "software_engineering"
  ],
  "LiveCodeBench - Medium": [
    "software_engineering"
  ],
  "LiveCodeBench - Overall": [
    "software_engineering"
  ],
  "LiveCodeBench Pro": [
    "software_engineering"
  ],
  "LiveCodeBench v5": [
    "software_engineering"
  ],
  "LiveCodeBench v5 24.12-25.2": [
    "software_engineering"
  ],
  "LiveCodeBench v6": [
    "software_engineering"
  ],
  "LiveCodeBench(01-09)": [
    "software_engineering"
  ],
  "Llm Stats": [
    "general"
  ],
  "LongBench v2": [
    "linguistic_core"
  ],
  "LongFact Concepts": [
    "hallucination"
  ],
  "LongFact Objects": [
    "hallucination"
  ],
  "LongVideoBench": [
    "multimodal"
  ],
  "MASK": [
    "linguistic_core"
  ],
  "MATH": [
    "mathematics"
  ],
  "MATH (CoT)": [
    "mathematics"
  ],
  "MATH-500": [
    "mathematics"
  ],
  "MATH-Mc": [
    "mathematics"
  ],
  "MAXIFE": [
    "multimodal"
  ],
  "MBPP ++ base version": [
    "software_engineering"
  ],
  "MBPP EvalPlus (base)": [
    "software_engineering"
  ],
  "MBPP pass@1": [
    "software_engineering"
  ],
  "MBPP+": [
    "software_engineering"
  ],
  "MCP Atlas": [
    "agentic"
  ],
  "MCP-Mark": [
    "agentic"
  ],
  "MCP-Universe": [
    "agentic"
  ],
  "MEGA MLQA": [
    "knowledge"
  ],
  "MEGA TyDi QA": [
    "knowledge"
  ],
  "MEGA UDPOS": [
    "linguistic_core"
  ],
  "MEGA XCOPA": [
    "commonsense_reasoning"
  ],
  "MEGA XStoryCloze": [
    "commonsense_reasoning"
  ],
  "MEWC": [
    "knowledge"
  ],
  "MGSM": [
    "mathematics"
  ],
  "MIABench": [
    "applied_reasoning"
  ],
  "MIMIC CXR": [
    "multimodal",
    "natural_sciences"
  ],
  "MLE-Bench Lite": [
    "software_engineering"
  ],
  "MLVU": [
    "multimodal"
  ],
  "MLVU-M": [
    "multimodal"
  ],
  "MM IF-Eval": [
    "multimodal"
  ],
  "MM-BrowserComp": [
    "agentic",
    "multimodal"
  ],
  "MM-ClawBench": [
    "agentic",
    "multimodal"
  ],
  "MM-MT-Bench": [
    "agentic",
    "multimodal"
  ],
  "MM-Mind2Web": [
    "agentic",
    "multimodal"
  ],
  "MMAU": [
    "multimodal"
  ],
  "MMAU Music": [
    "multimodal"
  ],
  "MMAU Sound": [
    "multimodal"
  ],
  "MMAU Speech": [
    "multimodal"
  ],
  "MMBench": [
    "multimodal"
  ],
  "MMBench-V1.1": [
    "multimodal"
  ],
  "MMBench-Video": [
    "multimodal"
  ],
  "MMBench_test": [
    "multimodal"
  ],
  "MME": [
    "multimodal"
  ],
  "MME-RealWorld": [
    "multimodal"
  ],
  "MMLU": [
    "general",
    "knowledge"
  ],
  "MMLU (CoT)": [
    "general",
    "knowledge"
  ],
  "MMLU Chat": [
    "general",
    "knowledge"
  ],
  "MMLU French": [
    "general",
    "knowledge"
  ],
  "MMLU-Base": [
    "general",
    "knowledge"
  ],
  "MMLU-Pro": [
    "general",
    "knowledge"
  ],
  "MMLU-Pro leaderboard submissions (TIGER-Lab)": [
    "general",
    "knowledge"
  ],
  "MMLU-ProX": [
    "general",
    "knowledge"
  ],
  "MMLU-Redux": [
    "general",
    "knowledge"
  ],
  "MMLU-STEM": [
    "general",
    "knowledge",
    "natural_sciences",
    "mathematics"
  ],
  "MMLU-redux-2.0": [
    "general",
    "knowledge"
  ],
  "MMLongBench-Doc": [
    "multimodal",
    "linguistic_core"
  ],
  "MMMLU": [
    "multimodal",
    "knowledge"
  ],
  "MMMU": [
    "multimodal",
    "knowledge"
  ],
  "MMMU (val)": [
    "multimodal",
    "knowledge"
  ],
  "MMMU (validation)": [
    "multimodal",
    "knowledge"
  ],
  "MMMUval": [
    "multimodal",
    "knowledge"
  ],
  "MMSearch": [
    "multimodal",
    "applied_reasoning"
  ],
  "MMSearch-Plus": [
    "multimodal",
    "applied_reasoning"
  ],
  "MMStar": [
    "multimodal",
    "applied_reasoning"
  ],
  "MMT-Bench": [
    "multimodal",
    "applied_reasoning"
  ],
  "MMVU": [
    "multimodal",
    "knowledge"
  ],
  "MMVet": [
    "multimodal",
    "applied_reasoning"
  ],
  "MMVetGPT4Turbo": [
    "multimodal",
    "applied_reasoning"
  ],
  "MRCR": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR 128K (2-needle)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR 128K (4-needle)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR 128K (8-needle)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR 1M": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR 1M (pointwise)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR 64K (2-needle)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR 64K (4-needle)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR 64K (8-needle)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR v2": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MRCR v2 (8-needle)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MS MARCO (TREC)": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MT-Bench": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MTVQA": [
    "multimodal",
    "applied_reasoning"
  ],
  "MVBench": [
    "multimodal",
    "applied_reasoning"
  ],
  "MathArena Apex": [
    "mathematics"
  ],
  "MathVerse-Mini": [
    "mathematics",
    "multimodal"
  ],
  "MathVision": [
    "mathematics",
    "multimodal"
  ],
  "MathVista": [
    "mathematics",
    "multimodal"
  ],
  "MathVista-Mini": [
    "mathematics",
    "multimodal"
  ],
  "MedCode - Overall": [
    "software_engineering",
    "natural_sciences"
  ],
  "MedQA": [
    "knowledge",
    "natural_sciences"
  ],
  "MedScribe - Overall": [
    "linguistic_core",
    "natural_sciences"
  ],
  "MedXpertQA": [
    "knowledge",
    "natural_sciences"
  ],
  "Meld": [
    "linguistic_core",
    "commonsense_reasoning"
  ],
  "MobileMiniWob++_SR": [
    "agentic"
  ],
  "Mortgage Tax": [
    "finance",
    "applied_reasoning"
  ],
  "MotionBench": [
    "multimodal",
    "applied_reasoning"
  ],
  "Mt Bench": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "MuSR": [
    "commonsense_reasoning",
    "applied_reasoning"
  ],
  "MuirBench": [
    "multimodal",
    "applied_reasoning"
  ],
  "Multi-Challenge": [
    "general"
  ],
  "Multi-IF": [
    "applied_reasoning"
  ],
  "Multi-SWE-Bench Leaderboard": [
    "software_engineering"
  ],
  "MultiLF": [
    "logical_reasoning",
    "applied_reasoning"
  ],
  "Multilingual MGSM (CoT)": [
    "mathematics",
    "linguistic_core"
  ],
  "Multilingual MMLU": [
    "knowledge",
    "linguistic_core"
  ],
  "Multipl E": [
    "software_engineering"
  ],
  "MusicCaps": [
    "multimodal"
  ],
  "NIH/Multi-needle": [
    "hallucination"
  ],
  "NL2Repo": [
    "software_engineering"
  ],
  "NMOS": [
    "multimodal"
  ],
  "NOVA-63": [
    "knowledge"
  ],
  "NQ": [
    "knowledge"
  ],
  "NarrativeQA": [
    "linguistic_core"
  ],
  "Natural Questions": [
    "knowledge"
  ],
  "Natural2Code": [
    "software_engineering"
  ],
  "NaturalQuestions": [
    "knowledge"
  ],
  "Nexus": [
    "agentic"
  ],
  "NoLiMa 128K": [
    "robustness"
  ],
  "NoLiMa 32K": [
    "robustness"
  ],
  "NoLiMa 64K": [
    "robustness"
  ],
  "Nuscene": [
    "multimodal"
  ],
  "OCRBench v2": [
    "multimodal"
  ],
  "OCRBench-V2 (en)": [
    "multimodal"
  ],
  "OCRBench-V2 (zh)": [
    "multimodal"
  ],
  "ODinW": [
    "multimodal"
  ],
  "OJBench": [
    "software_engineering"
  ],
  "OJBench (C++)": [
    "software_engineering"
  ],
  "OSWorld": [
    "agentic"
  ],
  "OSWorld Extended": [
    "agentic"
  ],
  "OSWorld Screenshot-only": [
    "agentic"
  ],
  "OSWorld-G": [
    "agentic"
  ],
  "OSWorld-Verified": [
    "agentic"
  ],
  "Objectron": [
    "multimodal"
  ],
  "OctoCodingBench": [
    "software_engineering"
  ],
  "OfficeQA Pro": [
    "agentic"
  ],
  "OlympiadBench": [
    "mathematics"
  ],
  "Omni-MATH": [
    "mathematics"
  ],
  "OmniBench": [
    "general"
  ],
  "OmniBench Music": [
    "multimodal"
  ],
  "OmniDocBench 1.5": [
    "multimodal"
  ],
  "OmniGAIA": [
    "applied_reasoning"
  ],
  "OmniMath": [
    "mathematics"
  ],
  "Online Mind2Web": [
    "agentic"
  ],
  "Open-rewrite": [
    "software_engineering"
  ],
  "OpenAI MMLU": [
    "knowledge"
  ],
  "OpenAI-MRCR: 2 needle 128k": [
    "hallucination"
  ],
  "OpenAI-MRCR: 2 needle 1M": [
    "hallucination"
  ],
  "OpenAI-MRCR: 2 needle 256k": [
    "hallucination"
  ],
  "OpenEval": [
    "general"
  ],
  "OpenRCA": [
    "applied_reasoning"
  ],
  "OpenbookQA": [
    "commonsense_reasoning"
  ],
  "Openeval": [
    "general"
  ],
  "PIQA": [
    "commonsense_reasoning"
  ],
  "PMC-VQA": [
    "multimodal"
  ],
  "POPE": [
    "hallucination"
  ],
  "PaperBench": [
    "knowledge",
    "applied_reasoning",
    "agentic"
  ],
  "PathMCQA": [
    "natural_sciences",
    "knowledge"
  ],
  "PerceptionTest": [
    "multimodal",
    "applied_reasoning"
  ],
  "PhiBench": [
    "knowledge",
    "applied_reasoning"
  ],
  "PhysicsFinals": [
    "natural_sciences",
    "mathematics"
  ],
  "PinchBench": [
    "multimodal",
    "applied_reasoning"
  ],
  "Piqa": [
    "commonsense_reasoning"
  ],
  "PointGrounding": [
    "multimodal"
  ],
  "PolyMATH": [
    "mathematics",
    "applied_reasoning"
  ],
  "PolyMath-en": [
    "mathematics",
    "applied_reasoning"
  ],
  "PopQA": [
    "knowledge"
  ],
  "ProofBench - Overall": [
    "logical_reasoning",
    "mathematics"
  ],
  "ProtocolQA": [
    "knowledge",
    "applied_reasoning"
  ],
  "QMSum": [
    "linguistic_core"
  ],
  "Qasper": [
    "linguistic_core"
  ],
  "QuAC": [
    "linguistic_core"
  ],
  "QwenWebBench": [
    "knowledge",
    "applied_reasoning"
  ],
  "RAFT": [
    "linguistic_core",
    "applied_reasoning"
  ],
  "RULER": [
    "linguistic_core",
    "robustness"
  ],
  "RULER 1000K": [
    "linguistic_core",
    "robustness"
  ],
  "RULER 128k": [
    "linguistic_core",
    "robustness"
  ],
  "RULER 2048K": [
    "linguistic_core",
    "robustness"
  ],
  "RULER 512K": [
    "linguistic_core",
    "robustness"
  ],
  "RULER 64k": [
    "linguistic_core",
    "robustness"
  ],
  "RealWorldQA": [
    "knowledge",
    "applied_reasoning"
  ],
  "RefCOCO-avg": [
    "multimodal"
  ],
  "RefSpatialBench": [
    "multimodal",
    "commonsense_reasoning"
  ],
  "RepoBench": [
    "software_engineering"
  ],
  "RepoQA": [
    "software_engineering"
  ],
  "Reward-Bench": [
    "applied_reasoning",
    "general"
  ],
  "RewardBench": [
    "applied_reasoning",
    "general"
  ],
  "RoboSpatialHome": [
    "agentic",
    "multimodal"
  ],
  "SAGE": [
    "knowledge",
    "applied_reasoning"
  ],
  "SAT Math": [
    "mathematics"
  ],
  "SIFO": [
    "knowledge",
    "applied_reasoning"
  ],
  "SIFO-Multiturn": [
    "knowledge",
    "applied_reasoning"
  ],
  "SQuALITY": [
    "linguistic_core"
  ],
  "STEM": [
    "natural_sciences",
    "mathematics"
  ],
  "SUNRGBD": [
    "multimodal"
  ],
  "SWE-Bench Multimodal": [
    "software_engineering",
    "multimodal"
  ],
  "SWE-Bench Pro": [
    "software_engineering",
    "agentic"
  ],
  "SWE-bench Verified Mini": [
    "software_engineering"
  ],
  "SWE-Lancer": [
    "software_engineering"
  ],
  "SWE-Lancer (IC-Diamond subset)": [
    "software_engineering"
  ],
  "SWE-Perf": [
    "software_engineering"
  ],
  "SWE-PolyBench": [
    "software_engineering",
    "applied_reasoning"
  ],
  "SWE-Review": [
    "software_engineering"
  ],
  "SWE-bench": [
    "software_engineering"
  ],
  "SWE-bench Multilingual": [
    "software_engineering"
  ],
  "SWE-bench Verified": [
    "software_engineering"
  ],
  "SWE-bench Verified (Agentic Coding)": [
    "software_engineering",
    "agentic"
  ],
  "SWE-bench Verified (Agentless)": [
    "software_engineering"
  ],
  "SWE-bench Verified (Multiple Attempts)": [
    "software_engineering"
  ],
  "SWT-Bench": [
    "software_engineering"
  ],
  "SciArena": [
    "natural_sciences",
    "applied_reasoning"
  ],
  "SciArena leaderboard API": [
    "natural_sciences",
    "applied_reasoning"
  ],
  "SciCode": [
    "natural_sciences",
    "software_engineering"
  ],
  "ScienceAgentBench": [
    "natural_sciences",
    "agentic"
  ],
  "ScienceQA": [
    "natural_sciences",
    "knowledge"
  ],
  "ScienceQA Visual": [
    "natural_sciences",
    "multimodal"
  ],
  "ScreenSpot": [
    "agentic",
    "multimodal"
  ],
  "ScreenSpot Pro": [
    "agentic",
    "multimodal"
  ],
  "Seal-0": [
    "safety"
  ],
  "SecCodeBench": [
    "software_engineering",
    "safety"
  ],
  "SimpleQA": [
    "knowledge",
    "hallucination"
  ],
  "SimpleSafetyTests": [
    "safety"
  ],
  "SimpleVQA": [
    "multimodal"
  ],
  "SkillsBench": [
    "applied_reasoning"
  ],
  "SlakeVQA": [
    "multimodal"
  ],
  "Social IQa": [
    "commonsense_reasoning"
  ],
  "Spider": [
    "software_engineering",
    "logical_reasoning"
  ],
  "SummScreenFD": [
    "linguistic_core"
  ],
  "SuperGLUE": [
    "linguistic_core",
    "general"
  ],
  "SuperGPQA": [
    "knowledge",
    "applied_reasoning"
  ],
  "TAU3-Bench": [
    "agentic"
  ],
  "TIR-Bench": [
    "agentic"
  ],
  "TLDR9+ (test)": [
    "linguistic_core"
  ],
  "Tau Bench": [
    "agentic"
  ],
  "Tau Bench Airline": [
    "agentic"
  ],
  "Tau2-Bench": [
    "agentic"
  ],
  "Tau2-Bench Airline": [
    "agentic"
  ],
  "Tau2-Bench Retail": [
    "agentic"
  ],
  "Tau2-Bench Telecom": [
    "agentic"
  ],
  "Tax Eval v2": [
    "finance",
    "applied_reasoning"
  ],
  "TempCompass": [
    "knowledge",
    "robustness"
  ],
  "Terminal Bench": [
    "agentic"
  ],
  "Terminal-Bench 2.0": [
    "agentic"
  ],
  "TerminalBench Hard": [
    "agentic"
  ],
  "Terminus": [
    "agentic"
  ],
  "TextVQA": [
    "multimodal"
  ],
  "TheoremQA": [
    "mathematics",
    "logical_reasoning"
  ],
  "Theory of Mind": [
    "commonsense_reasoning"
  ],
  "Toolathlon": [
    "agentic"
  ],
  "Translation Set1\u2192en COMET22": [
    "linguistic_core"
  ],
  "Translation Set1\u2192en spBleu": [
    "linguistic_core"
  ],
  "Translation en\u2192Set1 COMET22": [
    "linguistic_core"
  ],
  "Translation en\u2192Set1 spBleu": [
    "linguistic_core"
  ],
  "TriviaQA": [
    "knowledge"
  ],
  "TruthfulQA": [
    "hallucination"
  ],
  "TydiQA": [
    "linguistic_core"
  ],
  "USACO": [
    "mathematics",
    "logical_reasoning"
  ],
  "USAMO25": [
    "mathematics"
  ],
  "Uniform Bar Exam": [
    "law"
  ],
  "V*": [
    "general"
  ],
  "VATEX": [
    "multimodal"
  ],
  "VCR_en_easy": [
    "multimodal"
  ],
  "VIBE": [
    "agentic"
  ],
  "VIBE Android": [
    "agentic"
  ],
  "VIBE Backend": [
    "agentic"
  ],
  "VIBE Simulation": [
    "agentic"
  ],
  "VIBE Web": [
    "agentic"
  ],
  "VIBE iOS": [
    "agentic"
  ],
  "VIBE-Pro": [
    "agentic"
  ],
  "VITA-Bench": [
    "multimodal"
  ],
  "VLMsAreBlind": [
    "multimodal"
  ],
  "VQA-Rad": [
    "multimodal"
  ],
  "VQAv2": [
    "multimodal"
  ],
  "VQAv2 (test)": [
    "multimodal"
  ],
  "VQAv2 (val)": [
    "multimodal"
  ],
  "Vals AI": [
    "general"
  ],
  "Vals Index": [
    "general"
  ],
  "Vals Multimodal Index": [
    "multimodal"
  ],
  "Vending-Bench 2": [
    "agentic"
  ],
  "Vibe Code Bench - Overall": [
    "software_engineering"
  ],
  "Vibe-Eval": [
    "applied_reasoning",
    "multimodal"
  ],
  "Video-MME": [
    "multimodal"
  ],
  "Video-MME (long, no subtitles)": [
    "multimodal"
  ],
  "VideoMME w sub.": [
    "multimodal"
  ],
  "VideoMME w/o sub.": [
    "multimodal"
  ],
  "VideoMMMU": [
    "multimodal"
  ],
  "Virology Capabilities Test": [
    "natural_sciences"
  ],
  "Vision2Web": [
    "agentic"
  ],
  "VisuLogic": [
    "multimodal"
  ],
  "VisualWebBench": [
    "agentic"
  ],
  "VocalSound": [
    "multimodal"
  ],
  "VoiceBench Avg": [
    "linguistic_core"
  ],
  "WMDP": [
    "safety"
  ],
  "WMT 2014": [
    "linguistic_core"
  ],
  "WMT23": [
    "linguistic_core"
  ],
  "WMT24++": [
    "linguistic_core"
  ],
  "We-Math": [
    "mathematics"
  ],
  "WebVoyager": [
    "agentic"
  ],
  "WideSearch": [
    "agentic"
  ],
  "Wild Bench": [
    "general"
  ],
  "WildBench": [
    "general"
  ],
  "Winogrande": [
    "commonsense_reasoning"
  ],
  "Wordle Arena": [
    "logical_reasoning"
  ],
  "WorldVQA": [
    "multimodal"
  ],
  "WritingBench": [
    "linguistic_core"
  ],
  "XLSum English": [
    "linguistic_core"
  ],
  "XSum": [
    "linguistic_core"
  ],
  "Xstest": [
    "safety"
  ],
  "ZClawBench": [
    "law"
  ],
  "ZEROBench": [
    "general"
  ],
  "ZEROBench-Sub": [
    "general"
  ],
  "ZebraLogic": [
    "logical_reasoning"
  ]
}