JURY-RL decouples vote-based answer proposals from Lean verification for stable label-free RLVR, matching supervised pass@1 while improving generalization on math benchmarks.
- Preservation of constraints and boundary assumptions
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
JURY-RL: Votes Propose, Proofs Dispose for Label-Free RLVR
JURY-RL decouples vote-based answer proposals from Lean verification for stable label-free RLVR, matching supervised pass@1 while improving generalization on math benchmarks.