Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Preference Classification on Helpfulness & Harmlessness (HH) (test)
Loading...
70.4
Classification Accuracy
BTPO
62.392
64.471
66.55
68.629
Oct 17, 2025
Classification Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Classification Accuracy
BTPO
Backbone=Qwen2.5-7B-In...
2025.10
70.4
GRPO (pair)
Backbone=Qwen2.5-7B-In...
2025.10
64.7
RM-R1
Backbone=Qwen2.5-7B-In...
2025.10
62.7
RM-R1-finetuned
Backbone=Qwen2.5-7B-In...
2025.10
62.7
Feedback
Search any
task
Search any
task