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

Question Answering on TriviaQA (ΔLL (%))

18.5ΔLL (%)

CE

-2.668162.827428.32313.81858May 28, 2026
Updated 5d ago

Evaluation Results

MethodLinks
2026.05
18.5
2026.05
18.2
2026.05
16.4
2026.05
15.5
2026.05
15.2
2026.05
14.8
2026.05
12.8
2026.05
11.5
2026.05
1.2
2026.05
1.1
2026.05
0.9
2026.05
0.4
2026.05
0.2
2026.05
-0.3
2026.05
-0.4
2026.05
-0.5
2026.05
-0.952
2026.05
-1.854