Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Capability Elicitation on Olympiads (test)
Loading...
88
Elicitation Rate (U_final / U_benign)
SFT+RL
10
30.25
50.5
70.75
Apr 23, 2026
Elicitation Rate (U_final / U_benign)
Updated 3mo ago
Evaluation Results
Method
Method
Links
Elicitation Rate (U_final / U_benign)
SFT+RL
Supervisor=LLAMA 3.1 8...
2026.04
88
SFT+RL
Supervisor=GPT 4o mini...
2026.04
86
SFT
Supervisor=GPT 4o mini...
2026.04
85
SFT
Supervisor=LLAMA 3.1 8...
2026.04
40
RL
Supervisor=GPT 4o mini...
2026.04
34
RL
Supervisor=LLAMA 3.1 8...
2026.04
13
Feedback
Search any
task
Search any
task