Off-policy GRPO, which estimates advantages from a slightly stale policy, matches or improves on-policy GRPO on math reasoning tasks and is supported by a new policy-improvement lower bound.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Revisiting Group Relative Policy Optimization: Insights into On-Policy and Off-Policy Training
Off-policy GRPO, which estimates advantages from a slightly stale policy, matches or improves on-policy GRPO on math reasoning tasks and is supported by a new policy-improvement lower bound.