CounterRoute is an online reinforcement‑learning framework that jointly learns how to route a language model’s reasoning between a ‘think’ and a ‘direct answer’ mode, using a single shared policy derived from a dual‑mode checkpoint. It employs counterfactual rollouts to credit routing decisions and a curriculum that starts with forced dual‑mode rollouts before shifting to self‑routed updates, achieving better accuracy‑efficiency trade‑offs across nine benchmarks. The method reduces generated tokens by up to 51% on Qwen3‑8B while improving macro‑average accuracy, and its routing strategy generalizes to unseen coding, science, and commonsense tasks.
By Ruochen Jiao, Besnik Fetahu, Zhenyu Shi, Priyanka Nigam
arXiv:2608. 02087v2 Announce Type: replace Abstract: Post-training Large Language Models (LLMs) with Reinforcement Learning (RL) has become an important tool for improving model capabilities, but the LLM action-space structure introduces challenges distinct from classical RL, with implications for inducing exploration.
By Jim Dilkes, Vahid Yazdanpanah, Sebastian Stein
arXiv:2609.37362v1 Announce Type: new
Abstract: Large language model (LLM) routing aims to assign each query to the most suitable model from a heterogeneous candidate pool, improving the quality--eff...
By Guannan Lai, Han-Jia Ye
arXiv:2510. 18814v4 Announce Type: replace-cross Abstract: Can language models improve their reasoning performance without external rewards, using only their own sampled responses for training?
By Mengqi Li, Lei Zhao, Anthony Man-Cho So, Ruoyu Sun, Xiao Li
The paper introduces Test‑Time Policy Optimization (TTPO), an approach that enables large language models to improve mathematical reasoning without relying on ground‑truth labels. TTPO uses majority‑vote pseudo‑labels and an asymmetric objective: it distills rollouts that agree with the pseudo‑label via On‑Policy Self‑Distillation and penalizes disagreeing rollouts with Grouped Reinforcement Learning. Token‑level selection further refines the process, down‑weighting already‑converged positions during distillation and penalizing only confident errors during RL. Experiments show that TTPO matches label‑supervised OPSD on five competition‑level benchmarks, boosts Qwen3‑1.7B from 38.0 % to 45.2 % in test‑time training, and achieves significant gains without explicit reasoning steps, while also generalizing well across tasks.
By Aozhe Wang, Zhengxi Lu, Jianze Wang, Shangke Lv, Ying Liu, Weiming Lu, Jun Xiao, Yueting Zhuang, Hua Yang, Qianglong Chen, Yongliang Shen
arXiv:2605.10194v2 Announce Type: replace
Abstract: On-policy self-distillation (OPSD) uses a model as its own teacher under privileged context, providing token-level supervision on the model's own r...
By Jiaxuan Wang, Xuan Ouyang, Zhiyu Chen, Yulan Hu, Lan-Zhe Guo
arXiv:2608. 11573v1 Announce Type: cross Abstract: Achieving effective self-correction, where models verify and correct their own mistakes, remains a fundamental challenge for large language models (LLMs).
By Vu Duc Anh, Nhat M. Hoang, Do Xuan Long, Cong-Duy Nguyen, Ponhvoan Srey, Luu Anh Tuan
arXiv:2608. 06296v1 Announce Type: new Abstract: On-policy (Self-)Distillation (OPD / OPSD) has shown strong potential for post-training large language models (LLMs).
By Yijiang Li, Bingyang Wang, Yijun Liang, Yunjie Tian, Di Fu, Nuno Vasconcelos
arXiv:2609.35954v1 Announce Type: cross
Abstract: Large language model post-training generates self-generated rollouts through reinforcement learning and on-policy distillation, yet this experience i...
By Zhiwei Zhang, Huayu Deng, Fei Zhao, Jiayan Fu, Bin Liang, Kam-Fai Wong, Mu Chuan
arXiv:2608.23493v1 Announce Type: new
Abstract: Self-reflection is a powerful mechanism for credit assignment in human learning, converting sparse outcome feedback into actionable guidance. However,...
By Jialong Liu, Yuling Shi, Ning Yang, Xiaodong Gu, Zuchao Li
arXiv:2507.21931v2 Announce Type: replace-cross
Abstract: Large Language Models (LLMs) often produce plausible but poorly-calibrated answers, limiting their reliability on reasoning-intensive tasks....
By Carel van Niekerk, Renato Vukovic, Benjamin Ruppik, Hsien-chin Lin, Shutong Feng, Milica Ga\v{s}i\'c
Achieving effective self-correction, where models verify and correct their own mistakes, remains a fundamental challenge for large language models (LLMs). In this work, we propose Self-Fix Step-DPO (SFS-DPO), a reinforcement learning based, two-stage framework for step-level self-verification and self-correction.