Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Post-RL Evaluation on HealthBench Hard
Loading...
32.1
Accuracy
MRRG
25.34
27.095
28.85
30.605
Jul 2, 2026
Accuracy
Updated 23d ago
Evaluation Results
Method
Method
Links
Accuracy
MRRG
2026.07
32.1
SVRG
Sub-reward (SR)=true
2026.07
28.7
Chasing the Tail
2026.07
27
SVRG
Sub-reward (SR)=false
2026.07
26
Base policy
Backbone=Qwen2.5-3B-In...
2026.07
25.6
Feedback
Search any
task
Search any
task