Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Temporal Regret Control on CAP-GSM8K GPT Claude Frontier Models (E=100)
Loading...
2.3
RCE_temp
Trivium
-0.356
17.572
35.5
53.428
Jun 3, 2026
RCE_temp
Alpha-hat
Commit Rate
Verdict Margin
Updated 1mo ago
Evaluation Results
Method
Method
Links
RCE_temp
Alpha-hat
Commit Rate
Verdict Margin
Trivium
Model=Claude-Sonnet-4....
2026.06
2.3
0.5
100
8
Trivium
Model=GPT-4o, Horizon...
2026.06
3
0.65
100
6
Trivium
Model=GPT-3.5-Turbo, H...
2026.06
3.7
0.8
100
5
Reactive
Model=Claude-Sonnet-4....
2026.06
20
4.3
0
-
RLVR
Model=Claude-Sonnet-4....
2026.06
21.3
4.6
0
-
Reactive
Model=GPT-4o, Horizon...
2026.06
57.3
12.4
0
-
RLVR
Model=GPT-4o, Horizon...
2026.06
58
12.6
0
-
RLVR
Model=GPT-3.5-Turbo, H...
2026.06
68
14.8
0
-
Reactive
Model=GPT-3.5-Turbo, H...
2026.06
68.7
14.9
0
-
Feedback
Search any
task
Search any
task