ResearchTasksSemantic Similarity EvaluationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedInsurance Tasks N=1500 (test)DeepSeek-R1 + Fine-tune0.869Mean BERT Cosine Similarity6Feb 26, 2026