BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping
Paper
•
2510.18927
•
Published
•
82
•
3
Alexey Gorbatovski
Myashka
AI & ML interests
NLP Alignment
Recent Activity
commented on
a paper
17 days ago
BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via
Balanced Policy Optimization with Adaptive Clipping
new activity
28 days ago
agentica-org/DeepScaleR-Preview-Dataset:There are no answers for 6 samples
updated
a model
2 months ago
Myashka/Qwen2.5-7B-UltraChat200K_EMA_SFT-Lr_3e_6-Alpha_0.01
Organizations
None yet
commented
a
paper
17 days ago
There are no answers for 6 samples
#4 opened 28 days ago
by
Myashka
Librarian Bot: Add language metadata for dataset
#2 opened over 1 year ago
by
librarian-bot