AI RESEARCH
Dual Consensus: Escaping from Spurious Majority in Unsupervised RLVR via Two-Stage Vote Mechanism
arXiv CS.LG
•
ArXi:2603.16223v1 Announce Type: new Current label-free RLVR approaches for large language models (LLMs), such as TTRL and Self-reward, have nstrated effectiveness in improving the performance of LLMs on complex reasoning tasks. However, these methods rely heavily on accurate pseudo-label estimation and converge on spurious yet popular answers, thereby trapping in a dominant mode and limiting further improvements. Building on this, we propose Dual Consensus Reinforcement Learning (DCRL), a novel self-supervised