Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Semantic Textual Similarity on STS Benchmark (test)

0.919Pearson Correlation (r)

Prompt-based FT (hard) + PCP

0.2288560.4080280.58720.766372Mar 29, 2018Jul 28, 2019Nov 25, 2020Mar 26, 2022Jul 25, 2023Nov 22, 2024Mar 24, 2026
Updated 24d ago

Evaluation Results

MethodLinks
2023.05
0.919
2018.04
0.81
2018.04
0.809
2018.04
0.808
2018.04
0.784
2018.04
0.782
2018.03
0.782
2018.04
0.781
2018.04
0.758
2018.04
0.755
2018.04
0.731
2018.04
0.72
2018.03
0.719
2023.05
0.715
2018.04
0.649
2018.04
0.639
2026.03
0.6305
2026.03
0.6286
2026.03
0.6162
2026.03
0.5933
2026.03
0.5651
2026.03
0.5584
2026.03
0.5553
2026.03
0.5529
2026.03
0.542
2026.03
0.5353
2026.03
0.5337
2026.03
0.5297
2026.03
0.5285
2026.03
0.5087
2026.03
0.4953
2026.03
0.494
2026.03
0.4891
2026.03
0.4577
2026.03
0.4563
2026.03
0.4329
2026.03
0.4231
2026.03
0.4181
2026.03
0.4171
2026.03
0.4072
2026.03
0.4052
2026.03
0.3948
2026.03
0.3912
2026.03
0.3682
2026.03
0.3667
2026.03
0.2554