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

Sentence completion on HellaSwag (test)

72.35Accuracy

Coherence Boosting (GPT-3 175B)

Updated 1mo ago

Evaluation Results

Method	Links
Coherence Boosting (GPT-3 175B) 2021.10		72.35
GPT-3 175B 2021.10		62.66
GPT-3 175B 2021.10		59.18
Coherence Boosting (GPT-2 XL) 2021.10		47.66
GPT-2 XL (1.6B) 2021.10		42.6
GPT-2 XL (1.6B) 2021.10		40
PGM 6 / 6 (1024) 2025.05		34.27
PGM 8 / 8 2025.05		33.2
PGM 6 / 6 (1024) 2025.05		32.55
Coherence Boosting (GPT-2 Small) 2021.10		31.84
PGM 8 / 8 2025.05		31.62
MDLM 2025.05		31.36
GPT-2 Small (125M) 2021.10		30.99
MDLM 2025.05		30.75
GPT-2 Small (125M) 2021.10		28.92