mlfoundations-dev/hero_run_2_fix_conversations_eval_03-18-25_00-30-42_0981
收藏数据链接:
官方服务:
资源简介:
该数据集为预计算的模型输出,用于评估模型在LiveCodeBench、AIME24、AIME25、AMC23、GPQADiamond和MATH500六个不同任务上的表现。每个任务包含了多次运行的详细准确率、解决的问题数量和总问题数量。
This dataset contains precomputed model outputs for evaluating the performance of the model on six different tasks: LiveCodeBench, AIME24, AIME25, AMC23, GPQADiamond, and MATH500. Each task includes detailed accuracy, number of questions solved, and total number of questions across multiple runs.
提供机构:
mlfoundations-dev


