Peer-Predictive Self-Training for Language Model Reasoning

arXiv:2604.13356v1 Announce Type: cross Abstract: Mechanisms for continued self-improvement of language models without external supervision remain an open challenge. We propose Peer-Predictive Self-Training (PST), a label-free fine-tuning framework in which multiple language models improve collaboratively by leveraging a cross-model aggregated response as an internal training signal. Given a prompt question, the models generate responses sequentially; the final aggregated answer, often more reli