retraction / README.md
HoangHa's picture
Upload dataset
d3c213a verified
metadata
dataset_info:
  - config_name: 20260223_openai_gpt-5.2_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: reasoning_content
        dtype: 'null'
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 1976867
        num_examples: 4775
    download_size: 167829
    dataset_size: 1976867
  - config_name: 20260223_openrouter_claude-opus-4-6_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: 'null'
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: 'null'
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 690
        num_examples: 1
    download_size: 12721
    dataset_size: 690
  - config_name: 20260223_openrouter_moonshotai_Kimi-K2.5_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: reasoning_content
        dtype: string
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 38255
        num_examples: 10
    download_size: 36692
    dataset_size: 38255
  - config_name: 20260223_openrouter_zai-org_GLM-5_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: reasoning_content
        dtype: string
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 5866661
        num_examples: 680
    download_size: 2193043
    dataset_size: 5866661
  - config_name: 20260223_vertex_gemini-3-pro-preview_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: reasoning_content
        dtype: string
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 3661513
        num_examples: 4775
    download_size: 809959
    dataset_size: 3661513
  - config_name: 20260223_vllm_Qwen_Qwen3-8B_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: 'null'
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 1892663
        num_examples: 4775
    download_size: 157276
    dataset_size: 1892663
  - config_name: 20260223_vllm_Qwen_Qwen3.5-397B-A17B_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: string
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 59025499
        num_examples: 4775
    download_size: 18959923
    dataset_size: 59025499
  - config_name: 20260223_vllm_allenai_Olmo-3.1-32B-Think_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: string
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 17644903
        num_examples: 4775
    download_size: 5941064
    dataset_size: 17644903
  - config_name: 20260223_vllm_google_gemma-3-27b-it_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: 'null'
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 1978412
        num_examples: 4775
    download_size: 159648
    dataset_size: 1978412
  - config_name: 20260223_vllm_meta-llama_Llama-3.1-8B-Instruct_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: 'null'
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 1989391
        num_examples: 4775
    download_size: 159660
    dataset_size: 1989391
  - config_name: 20260223_vllm_meta-llama_Llama-3.3-70B-Instruct_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: 'null'
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 2005153
        num_examples: 4775
    download_size: 158664
    dataset_size: 2005153
  - config_name: 20260223_vllm_mistralai_Ministral-3-14B-Reasoning-2512_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: 'null'
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 2103404
        num_examples: 4775
    download_size: 159211
    dataset_size: 2103404
  - config_name: 20260223_vllm_nvidia_NVIDIA-Nemotron-3-Nano-30B-A3B-BF16_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: string
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 8558357
        num_examples: 4775
    download_size: 1939743
    dataset_size: 8558357
  - config_name: 20260223_vllm_openai_gpt-oss-120b_A-1_noweb
    features:
      - name: paper_id
        dtype: string
      - name: citation_count
        dtype: int64
      - name: is_retracted
        dtype: bool
      - name: question_id
        dtype: string
      - name: question_text
        dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        dtype: string
      - name: classification
        dtype: string
      - name: input_tokens
        dtype: 'null'
      - name: output_tokens
        dtype: 'null'
      - name: reasoning_content
        dtype: 'null'
      - name: repetition
        dtype: int64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: temperature
        dtype: 'null'
      - name: finish_reason
        dtype: 'null'
      - name: elapsed_seconds
        dtype: 'null'
      - name: run_id
        dtype: 'null'
      - name: timestamp
        dtype: 'null'
    splits:
      - name: train
        num_bytes: 5107186
        num_examples: 4775
    download_size: 989006
    dataset_size: 5107186
  - config_name: 20260303_vllm_Qwen_Qwen3.5-9B-Base_A-1_noweb
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: string
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 1135101722
        num_examples: 4775
    download_size: 1078988432
    dataset_size: 1135101722
  - config_name: 20260303_vllm_Qwen_Qwen3.5-9B_A-1_noweb
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: string
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 908358000
        num_examples: 4775
    download_size: 858165240
    dataset_size: 908358000
  - config_name: 20260303_vllm_Qwen_Qwen3.5-9B_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: string
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: float64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 896110782
        num_examples: 4775
    download_size: 846756024
    dataset_size: 896110782
  - config_name: 20260304_vllm_Qwen_Qwen3-8B_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: string
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 904345106
        num_examples: 4775
    download_size: 854776361
    dataset_size: 904345106
  - config_name: 20260304_vllm_Qwen_Qwen3.5-9B-Base_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: string
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 286976554
        num_examples: 4775
    download_size: 269833222
    dataset_size: 286976554
  - config_name: 20260304_vllm_google_gemma-3-27b-it_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: 'null'
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 37763540
        num_examples: 4775
    download_size: 35531179
    dataset_size: 37763540
  - config_name: 20260304_vllm_meta-llama_Llama-3.1-8B-Instruct_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: 'null'
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 34603072
        num_examples: 4775
    download_size: 32431511
    dataset_size: 34603072
  - config_name: 20260304_vllm_meta-llama_Llama-3.3-70B-Instruct_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: 'null'
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 34491417
        num_examples: 4775
    download_size: 32349606
    dataset_size: 34491417
  - config_name: 20260304_vllm_mistralai_Ministral-3-14B-Reasoning-2512_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: 'null'
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 37601488
        num_examples: 4775
    download_size: 35276126
    dataset_size: 37601488
  - config_name: 20260304_vllm_nvidia_NVIDIA-Nemotron-3-Nano-30B-A3B-BF16_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: string
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 272369661
        num_examples: 4775
    download_size: 258754243
    dataset_size: 272369661
  - config_name: 20260304_vllm_openai_gpt-oss-120b_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: 'null'
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 77169283
        num_examples: 4775
    download_size: 73410616
    dataset_size: 77169283
  - config_name: 20260305_openai_gpt-5.2_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: string
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: float64
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 4739031
        num_examples: 884
    download_size: 4402072
    dataset_size: 4739031
  - config_name: 20260305_vertexai_publishers_google_models_gemini-3.1-pro-preview_all_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: 'null'
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 30072531
        num_examples: 4775
    download_size: 27994039
    dataset_size: 30072531
  - config_name: 20260305_vllm_allenai_Olmo-3.1-32B-Think_A-1_web
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        list:
          - name: title
            dtype: string
          - name: snippet
            dtype: string
          - name: link
            dtype: string
          - name: full_content
            dtype: string
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: string
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        list:
          - name: token
            dtype: string
          - name: logprob
            dtype: float64
          - name: top_k
            list:
              - name: token
                dtype: string
              - name: logprob
                dtype: float64
      - name: mean_logprob
        dtype: float64
      - name: min_logprob
        dtype: float64
      - name: mean_entropy
        dtype: float64
      - name: max_entropy
        dtype: float64
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 582564789
        num_examples: 4775
    download_size: 547657023
    dataset_size: 582564789
  - config_name: >-
      20260309_vertexai_publishers_google_models_gemini-3.1-pro-preview_all_noweb
    features:
      - name: run_id
        dtype: string
      - name: paper_id
        dtype: string
      - name: question_id
        dtype: string
      - name: repetition
        dtype: int64
      - name: timestamp
        dtype: string
      - name: is_retracted
        dtype: bool
      - name: citation_count
        dtype: int64
      - name: question_text
        dtype: string
      - name: web_search_context
        dtype: 'null'
      - name: response_text
        dtype: string
      - name: response_json
        struct:
          - name: paper_is_retracted
            dtype: bool
      - name: reasoning_content
        dtype: 'null'
      - name: classification
        dtype: string
      - name: finish_reason
        dtype: string
      - name: logprobs
        dtype: 'null'
      - name: mean_logprob
        dtype: 'null'
      - name: min_logprob
        dtype: 'null'
      - name: mean_entropy
        dtype: 'null'
      - name: max_entropy
        dtype: 'null'
      - name: input_tokens
        dtype: int64
      - name: output_tokens
        dtype: int64
      - name: elapsed_seconds
        dtype: 'null'
      - name: model
        dtype: string
      - name: provider
        dtype: string
      - name: temperature
        dtype: float64
    splits:
      - name: train
        num_bytes: 2336298
        num_examples: 4775
    download_size: 230956
    dataset_size: 2336298
configs:
  - config_name: 20260223_openai_gpt-5.2_A-1_noweb
    data_files:
      - split: train
        path: 20260223_openai_gpt-5.2_A-1_noweb/train-*
  - config_name: 20260223_openrouter_claude-opus-4-6_A-1_noweb
    data_files:
      - split: train
        path: 20260223_openrouter_claude-opus-4-6_A-1_noweb/train-*
  - config_name: 20260223_openrouter_moonshotai_Kimi-K2.5_A-1_noweb
    data_files:
      - split: train
        path: 20260223_openrouter_moonshotai_Kimi-K2.5_A-1_noweb/train-*
  - config_name: 20260223_openrouter_zai-org_GLM-5_A-1_noweb
    data_files:
      - split: train
        path: 20260223_openrouter_zai-org_GLM-5_A-1_noweb/train-*
  - config_name: 20260223_vertex_gemini-3-pro-preview_A-1_noweb
    data_files:
      - split: train
        path: 20260223_vertex_gemini-3-pro-preview_A-1_noweb/train-*
  - config_name: 20260223_vllm_Qwen_Qwen3-8B_A-1_noweb
    data_files:
      - split: train
        path: 20260223_vllm_Qwen_Qwen3-8B_A-1_noweb/train-*
  - config_name: 20260223_vllm_Qwen_Qwen3.5-397B-A17B_A-1_noweb
    data_files:
      - split: train
        path: 20260223_vllm_Qwen_Qwen3.5-397B-A17B_A-1_noweb/train-*
  - config_name: 20260223_vllm_allenai_Olmo-3.1-32B-Think_A-1_noweb
    data_files:
      - split: train
        path: 20260223_vllm_allenai_Olmo-3.1-32B-Think_A-1_noweb/train-*
  - config_name: 20260223_vllm_google_gemma-3-27b-it_A-1_noweb
    data_files:
      - split: train
        path: 20260223_vllm_google_gemma-3-27b-it_A-1_noweb/train-*
  - config_name: 20260223_vllm_meta-llama_Llama-3.1-8B-Instruct_A-1_noweb
    data_files:
      - split: train
        path: 20260223_vllm_meta-llama_Llama-3.1-8B-Instruct_A-1_noweb/train-*
  - config_name: 20260223_vllm_meta-llama_Llama-3.3-70B-Instruct_A-1_noweb
    data_files:
      - split: train
        path: 20260223_vllm_meta-llama_Llama-3.3-70B-Instruct_A-1_noweb/train-*
  - config_name: 20260223_vllm_mistralai_Ministral-3-14B-Reasoning-2512_A-1_noweb
    data_files:
      - split: train
        path: >-
          20260223_vllm_mistralai_Ministral-3-14B-Reasoning-2512_A-1_noweb/train-*
  - config_name: 20260223_vllm_nvidia_NVIDIA-Nemotron-3-Nano-30B-A3B-BF16_A-1_noweb
    data_files:
      - split: train
        path: >-
          20260223_vllm_nvidia_NVIDIA-Nemotron-3-Nano-30B-A3B-BF16_A-1_noweb/train-*
  - config_name: 20260223_vllm_openai_gpt-oss-120b_A-1_noweb
    data_files:
      - split: train
        path: 20260223_vllm_openai_gpt-oss-120b_A-1_noweb/train-*
  - config_name: 20260303_vllm_Qwen_Qwen3.5-9B-Base_A-1_noweb
    data_files:
      - split: train
        path: 20260303_vllm_Qwen_Qwen3.5-9B-Base_A-1_noweb/train-*
  - config_name: 20260303_vllm_Qwen_Qwen3.5-9B_A-1_noweb
    data_files:
      - split: train
        path: 20260303_vllm_Qwen_Qwen3.5-9B_A-1_noweb/train-*
  - config_name: 20260303_vllm_Qwen_Qwen3.5-9B_A-1_web
    data_files:
      - split: train
        path: 20260303_vllm_Qwen_Qwen3.5-9B_A-1_web/train-*
  - config_name: 20260304_vllm_Qwen_Qwen3-8B_A-1_web
    data_files:
      - split: train
        path: 20260304_vllm_Qwen_Qwen3-8B_A-1_web/train-*
  - config_name: 20260304_vllm_Qwen_Qwen3.5-9B-Base_A-1_web
    data_files:
      - split: train
        path: 20260304_vllm_Qwen_Qwen3.5-9B-Base_A-1_web/train-*
  - config_name: 20260304_vllm_google_gemma-3-27b-it_A-1_web
    data_files:
      - split: train
        path: 20260304_vllm_google_gemma-3-27b-it_A-1_web/train-*
  - config_name: 20260304_vllm_meta-llama_Llama-3.1-8B-Instruct_A-1_web
    data_files:
      - split: train
        path: 20260304_vllm_meta-llama_Llama-3.1-8B-Instruct_A-1_web/train-*
  - config_name: 20260304_vllm_meta-llama_Llama-3.3-70B-Instruct_A-1_web
    data_files:
      - split: train
        path: 20260304_vllm_meta-llama_Llama-3.3-70B-Instruct_A-1_web/train-*
  - config_name: 20260304_vllm_mistralai_Ministral-3-14B-Reasoning-2512_A-1_web
    data_files:
      - split: train
        path: 20260304_vllm_mistralai_Ministral-3-14B-Reasoning-2512_A-1_web/train-*
  - config_name: 20260304_vllm_nvidia_NVIDIA-Nemotron-3-Nano-30B-A3B-BF16_A-1_web
    data_files:
      - split: train
        path: >-
          20260304_vllm_nvidia_NVIDIA-Nemotron-3-Nano-30B-A3B-BF16_A-1_web/train-*
  - config_name: 20260304_vllm_openai_gpt-oss-120b_A-1_web
    data_files:
      - split: train
        path: 20260304_vllm_openai_gpt-oss-120b_A-1_web/train-*
  - config_name: 20260305_openai_gpt-5.2_A-1_web
    data_files:
      - split: train
        path: 20260305_openai_gpt-5.2_A-1_web/train-*
  - config_name: 20260305_vertexai_publishers_google_models_gemini-3.1-pro-preview_all_web
    data_files:
      - split: train
        path: >-
          20260305_vertexai_publishers_google_models_gemini-3.1-pro-preview_all_web/train-*
  - config_name: 20260305_vllm_allenai_Olmo-3.1-32B-Think_A-1_web
    data_files:
      - split: train
        path: 20260305_vllm_allenai_Olmo-3.1-32B-Think_A-1_web/train-*
  - config_name: >-
      20260306_vertexai_publishers_google_models_gemini-3.1-pro-preview_all_noweb
    data_files:
      - split: train
        path: >-
          20260306_vertexai_publishers_google_models_gemini-3.1-pro-preview_all_noweb/train-*
  - config_name: >-
      20260309_vertexai_publishers_google_models_gemini-3.1-pro-preview_all_noweb
    data_files:
      - split: train
        path: >-
          20260309_vertexai_publishers_google_models_gemini-3.1-pro-preview_all_noweb/train-*
  - config_name: evaluati_openrouter_Qwen_Qwen3-8B_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_Qwen_Qwen3-8B_A-1_noweb/train-*
  - config_name: evaluati_openrouter_Qwen_Qwen3.5-397B-A17B_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_Qwen_Qwen3.5-397B-A17B_A-1_noweb/train-*
  - config_name: evaluati_openrouter_allenai_Olmo-3.1-32B-Think_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_allenai_Olmo-3.1-32B-Think_A-1_noweb/train-*
  - config_name: evaluati_openrouter_claude-opus-4-6_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_claude-opus-4-6_A-1_noweb/train-*
  - config_name: evaluati_openrouter_gemini-3-pro-preview_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_gemini-3-pro-preview_A-1_noweb/train-*
  - config_name: evaluati_openrouter_google_gemma-3-27b-it_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_google_gemma-3-27b-it_A-1_noweb/train-*
  - config_name: evaluati_openrouter_meta-llama_Llama-3.1-8B-Instruct_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_meta-llama_Llama-3.1-8B-Instruct_A-1_noweb/train-*
  - config_name: evaluati_openrouter_meta-llama_Llama-3.3-70B-Instruct_A-1_noweb
    data_files:
      - split: train
        path: >-
          evaluati_openrouter_meta-llama_Llama-3.3-70B-Instruct_A-1_noweb/train-*
  - config_name: evaluati_openrouter_mistralai_Ministral-3-14B-Reasoning-2512_A-1_noweb
    data_files:
      - split: train
        path: >-
          evaluati_openrouter_mistralai_Ministral-3-14B-Reasoning-2512_A-1_noweb/train-*
  - config_name: evaluati_openrouter_moonshotai_Kimi-K2.5_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_moonshotai_Kimi-K2.5_A-1_noweb/train-*
  - config_name: evaluati_openrouter_nvidia_NVIDIA-Nemotron-3-Nano-30B-A3B-BF16_A-1_noweb
    data_files:
      - split: train
        path: >-
          evaluati_openrouter_nvidia_NVIDIA-Nemotron-3-Nano-30B-A3B-BF16_A-1_noweb/train-*
  - config_name: evaluati_openrouter_openai_gpt-5.2_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_openai_gpt-5.2_A-1_noweb/train-*
  - config_name: evaluati_openrouter_openai_gpt-oss-120b_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_openai_gpt-oss-120b_A-1_noweb/train-*
  - config_name: evaluati_openrouter_zai-org_GLM-5_A-1_noweb
    data_files:
      - split: train
        path: evaluati_openrouter_zai-org_GLM-5_A-1_noweb/train-*