Rubric Curriculum RL (RcRL) is a reinforcement learning approach that leverages the generation‑verification gap in non‑verifiable domains by using pairwise preferences over rollouts and a curriculum over rubric criteria. Unlike traditional methods that rely on absolute scoring and stationary judges, RcRL provides a more discriminative, non‑stationary signal, allowing policies to maintain output entropy and continue improving beyond 1,000 training steps. In experiments on creative writing and HealthBench, RcRL outperforms baseline models, achieving a 70.5% human win rate and significant accuracy gains over the base model.
By Sumeet Ramesh Motwani, Tejas Krishnan, Alistair Letcher, Charles London, Suhaas Bhat, Huitian Jiao, Philip Torr, Riashat Islam, Christopher Summerfield, Jakob Foerster, Christian Schroeder de Witt, Qilong Gu, Shital Shah
SERL‑SQL introduces a selective execution‑grounded reinforcement learning framework for multi‑turn Text‑to‑SQL agents. It samples on‑policy interaction trajectories and employs a training‑only teacher to re‑score student actions using execution feedback, converting the teacher‑student likelihood gap into bounded, masked weights that reweight GRPO advantages only on SQL and tool‑action tokens. Experiments on BIRD, Spider, and cross‑domain benchmarks show competitive performance, achieving 76.56% execution accuracy on BIRD‑Dev and 89.92% on Spider‑Test, and the reward‑based selection strategy approaches the oracle Best‑of‑N upper bound.
By Tao Liu, Tao Feng, Xiangheng Li, Jinwang Song, Yifan Li, Xiaoqing Cheng, Dixuan Zhang, Siquan Li, Lin Lan, Hongying Zan, Kunli Zhang, Chao Wu
The paper introduces MENTOR, a multi‑agent framework that provides on‑policy pedagogical intervention to address zero‑reward collapse in reinforcement learning from verifiable rewards. MENTOR diagnoses failures in real time and generates targeted prompts to guide the learning agent, integrating seamlessly with existing policy optimization pipelines such as GRPO. Experiments on math and code generation tasks show that MENTOR significantly reduces collapse, achieving up to 78.3% relative gains on Llama‑3.2‑3B‑Instruct.
By Guanlan Yue, Hang Yu, Xiaoyong Huang, Ruipeng Li, Wei Zhang, Bingchang Liu, Rui Wang
RL-Teacher is an open-source implementation of our interface to train AIs via occasional human feedback rather than hand-crafted reward functions. The underlying technique was developed as a step towards safe AI systems, but also applies to reinforcement learning problems with rewards that are hard to specify.
arXiv:2606. 24839v1 Announce Type: new Abstract: Agentic data analysis systems produce rich outputs, including code, numerical results, and verbal diagnostics.
By Tian Zheng, Kai-Tai Hsu
HarnessSQL introduces a harness‑native post‑training framework that keeps the full stateful, multi‑turn interaction structure of real‑world database agents during both supervised fine‑tuning and reinforcement learning. By building isolated executable database environments with hidden execution oracles and retaining only verified trajectories for full‑sequence SFT, the method enables execution‑reward RL that aligns training with inference. Experiments on Spider 2.0‑SQLite show dramatic gains in execution accuracy for compact models, and the approach transfers effectively to out‑of‑distribution interactive benchmarks such as BIRD‑Interact and LiveSQLBench.
By Haolin Yang, Jipeng Zhang, Jian Xie, Shuaishuai Gong, Sirui Han, Yike Guo
Reward models (RMs) provide critical feedback signals for LLM post-training, notably in reinforced fine-tuning (RFT) and reinforcement learning (RL) pipelines. However, current reward evaluation relies on heterogeneous criteria such as rule-based verifiers, ground-truth references, procedural checklists, and complex rubrics, where a unified mechanism to integrate all types of evidence remains unexplored.
The paper introduces Evidence‑Diagnosed Intervention Training (EDIT), a two‑phase framework designed to improve rubric‑faithful grading by large language models. EDIT‑SFT first identifies problematic reasoning steps using internal model signals—posterior belief over the final mark and input‑grounding scores—and revises only those steps with rubric checklists. EDIT‑RL then calibrates the grader with belief‑guided reward shaping, penalising harmful belief drifts while encouraging useful exploration. Experiments on two real‑world, multi‑subject grading benchmarks show that EDIT consistently outperforms strong supervised fine‑tuning and reinforcement learning baselines, with ablation studies confirming the importance of internal‑state diagnostics.
By Zhihao Wu, Linhai Zhang, Taiyi Wang, Runcong Zhao, Peter Andrews, Cesare Aloisi, Yulan He
arXiv:2606. 03980v1 Announce Type: new Abstract: Reward models (RMs) provide critical feedback signals for LLM post-training, notably in reinforced fine-tuning (RFT) and reinforcement learning (RL) pipelines.
By Tao Chen, Gangwei Jiang, Pengyu Cheng, Siyuan Huang, Yihao Liu, Jingwei Ni, Jiaqi Guo, Mengyu Zhou, Kai Tang, Junling Liu, Qinliang Su, Xiaoxi Jiang, Guanjun Jiang
arXiv:2610.23234v1 Announce Type: new
Abstract: In a hierarchical multi-agent system, a Leader assigns tasks and subagents carry them out. Jointly training the Leader and the subagents is difficult b...
By Yuhan Sun, Jiaxin Zhao, Yongdong Luo, Yibo Hu, Meiguang Jin, Junfeng Ma, Weixun Wang, Weihang Pan, Zhongan Bi
The paper introduces Recursive Reward Improvement (RRI), a method that enhances recursive self‑improvement loops by using an adversary to expose and fix weaknesses in model‑generated rubrics for open‑ended tasks. In RRI, an adversary crafts a rubric‑gaming answer, an honest answer is generated without rubric exposure, and up to two repair steps adjust the rubric so the honest answer scores higher. Experiments on HealthBench Professional show that RRI boosts adjusted scores by up to 12.4 points on a 27B model and yields significant gains on difficult and writing tasks, while also improving performance on unseen benchmarks PRBench Hard and ProfBench.
By Wei Dai, Qianchu Liu, Guanghui Qin, Xing Han, Tristan Naumann, Paul Pu Liang, Sheng Zhang
arXiv:2606. 09380v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has become a leading paradigm for improving the reasoning ability of large language models through outcome-based supervision.
By Han Zhou, Adam X. Yang, Laurence Aitchison, Anna Korhonen, Albert Q. Jiang