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

Conclusion Generation on EntailmentBank (test)

67BLEU

T5 original (large)

4.620.83753.2Jan 31, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
670.970.5
2026.01
650.970.45
2026.01
650.970.48
2026.01
640.970.44
2026.01
630.970.45
2026.01
630.970.46
2026.01
610.970.39
2026.01
600.970.46
2026.01
600.960.42
2026.01
590.960.34
2026.01
570.960.33
2026.01
570.960.31
2026.01
550.960.3
2026.01
540.960.22
2026.01
530.960.25
2026.01
460.960.23
2026.01
420.92-0.07
2026.01
390.91-0.13
2026.01
350.91-0.15
2026.01
350.91-0.2
2026.01
280.91-0.9
2026.01
270.9-0.97
2026.01
260.8-1.11
2026.01
250.79-1.14
2026.01
90.74-1.17
2026.01
70.74-1.2