A unified Pair-GRPO framework extends GRPO with soft and hard pairwise preference variants, proving gradient equivalence under Taylor expansion and delivering improved stability and performance in RLHF.
arXiv preprint arXiv:2405.10015 , year=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Unified Pair-GRPO Family: From Implicit to Explicit Preference Constraints for Stable and General RL Alignment
A unified Pair-GRPO framework extends GRPO with soft and hard pairwise preference variants, proving gradient equivalence under Taylor expansion and delivering improved stability and performance in RLHF.