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

Self-attention inverse temperature scaling analysis on OpenWebText (OWT)

5.7Tie (%)

nanoGPT

3.9324.3914.855.309May 12, 2026
Updated 20d ago

Evaluation Results

MethodLinks
2026.05
5.70.420.690.970.66
2026.05
40.460.571.030.6