ResearchDatasetsPRBenchFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsLegal ReasoningPRBench Legal-Hard0.377GPT-5 Grader Score4Legal ReasoningPRBench Legal0.482GPT-5 Grader Score4Legal reasoningPRBench-Legal 500 prompts44.1GPT-5 Grader Score4Generative PerformancePRBench0.8Pearson r1