Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Truthful Question Answering on TruthfulQA (Mean per-step regret)
Loading...
0.138
Mean Per-Step Regret
ϵ-FTRL
0.13528
0.15364
0.172
0.19036
Feb 23, 2026
Mean Per-Step Regret
Updated 1mo ago
Evaluation Results
Method
Method
Links
Mean Per-Step Regret
ϵ-FTRL
Strategy Category=Non-...
2026.02
0.138
Simpl
Strategy Category=Stat...
2026.02
0.141
LinUCB+KL
Strategy Category=Cont...
2026.02
0.15
No-Rewrite (NoRw)
Strategy Category=Base
2026.02
0.151
TS
Strategy Category=Non-...
2026.02
0.155
TS
Strategy Category=Cont...
2026.02
0.155
Para
Strategy Category=Stat...
2026.02
0.159
LinUCB
Strategy Category=Cont...
2026.02
0.161
Disamb
Strategy Category=Stat...
2026.02
0.166
FTPL
Strategy Category=Non-...
2026.02
0.167
LinEXP3
Strategy Category=Cont...
2026.02
0.171
EXP3
Strategy Category=Non-...
2026.02
0.173
Clarify
Strategy Category=Stat...
2026.02
0.18
LinFTPL
Strategy Category=Cont...
2026.02
0.18
Expand
Strategy Category=Stat...
2026.02
0.206
Feedback
Search any
task
Search any
task