ResearchTasksSemantic Similarity EvaluationSemantic Similarity EvaluationFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedInsurance Tasks N=1500 (test)DeepSeek-R1 + Fine-tune0.869Mean BERT Cosine Similarity6Feb 26, 2026