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

Language Modeling on PubMedQA MdQ

0PPL Change (%) vs Baseline

GAIN

-0.9565.49711.9518.403Apr 6, 2026
Updated 11d ago

Evaluation Results

MethodLinks
2026.04
0
2026.04
0.1
2026.04
0.2
2026.04
0.2
2026.04
0.3
2026.04
3.1
2026.04
3.7
2026.04
3.9
2026.04
4.2
2026.04
4.4
2026.04
4.5
2026.04
4.6
2026.04
6.8
2026.04
7.4
2026.04
7.5
2026.04
7.8
2026.04
8
2026.04
8.1
2026.04
8.7
2026.04
15
2026.04
16.2
2026.04
17.9
2026.04
19.3
2026.04
23.9