List your company

Natural Language Inference on EntailmentBank (test)

54BLEUllama3-70b-8192
41.5244.764851.24Jan 31, 2026

Evaluation results

MethodLinks
llama3-70b-8192Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=102026.01540.9754
Flan-T5 (xxl)Architecture=encoder-decoder, Inference Type Injection (INJ)=Yes, Num examples=52026.01530.9743
GPT-4-0613Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=102026.01530.9750
GPT-3.5-turbo-0125Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=102026.01520.9640
GPT-4-0613Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=52026.01520.9747
GPT-4-0613Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=02026.01520.9750
llama3-70b-8192Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=52026.01520.9752
CoT-T5 (11b)Architecture=encoder-decoder, Inference Type Injection (INJ)=Yes, Num examples=102026.01510.9739
CoT-T5 (11b)Architecture=encoder-decoder, Inference Type Injection (INJ)=Yes, Num examples=52026.01510.9739
Flan-T5 (xxl)Architecture=encoder-decoder, Inference Type Injection (INJ)=Yes, Num examples=102026.01510.9741
llama3-70b-8192Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=02026.01510.9747
CoT-T5 (11b)Architecture=encoder-decoder, Inference Type Injection (INJ)=Yes, Num examples=02026.01500.9736
Flan-T5 (xxl)Architecture=encoder-decoder, Inference Type Injection (INJ)=Yes, Num examples=02026.01500.9637
Flan-T5 (xxl)Architecture=encoder-decoder, Inference Type Injection (INJ)=No, Num examples=02026.01480.9636
GPT-3.5-turbo-0125Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=52026.01480.9635
GPT-4-0613Architecture=decoder only, Inference Type Injection (INJ)=No, Num examples=02026.01470.9640
CoT-T5 (11b)Architecture=encoder-decoder, Inference Type Injection (INJ)=No, Num examples=02026.01460.9631
GPT-3.5-turbo-0125Architecture=decoder only, Inference Type Injection (INJ)=Yes, Num examples=02026.01460.9631
llama3-70b-8192Architecture=decoder only, Inference Type Injection (INJ)=No, Num examples=02026.01440.9640
GPT-3.5-turbo-0125Architecture=decoder only, Inference Type Injection (INJ)=No, Num examples=02026.01420.9633