MAGIC introduces a dense‑reward reinforcement learning framework for generating mixed‑granularity agent graphs in large‑language‑model based multi‑agent systems. The method sequentially selects functional roles, instantiates them as either single agents or reusable groups, and connects them to existing units, optimizing the construction policy with intermediate feedback from probe‑based utility and structural signals. Experiments show MAGIC outperforms state‑of‑the‑art baselines on eight benchmarks and achieves strong inference efficiency.
By Kairui Yang, Ziheng Yi, Xunkai Li, Minghao An, Zhanke Liu, Zekai Chen, Rong-Hua Li
The paper investigates how entropy over chain‑of‑thought tokens influences policy decisions such as gradient application, pruning, and collapse detection. By separating scaffold tokens from substantive content, the authors analyze entropy, Kullback–Leibler divergence, and entropy velocity for each channel, proving differences between raw and content conventions and bounding answer diversity. Empirical results across 23 configurations show that scaffold tokens can account for up to 41% of high‑entropy positions, with entropy share growing through distillation, while content conventions outperform raw surprisal on compression tasks and reveal significant answer leakage in re‑fed chains.
By Marios Papamichalis, Regina Ruane
Matryoshka Attribution (MAttr) is a mask‑learning method that identifies nested subsets of a language model’s internal components by minimizing downstream loss. It uses a differentiable sigmoid top‑k operator and randomizes sparsity during training to produce an attribution ordering of components. MAttr tops the Mechanistic Interpretability Benchmark leaderboard and can be applied via reinforcement learning to pinpoint weight changes that control behaviors such as refusal in Llama 3.1 8B Instruct, where restoring just 1% of weights removes refusals while preserving capabilities.
By Aryaman Arora, Kirill Acharya, Nathan Hu, Yanzhe Zhang, Noah Goodman, Dan Jurafsky, Christopher Potts
Ladders-of-Thought (LoT) is a framework that enhances reasoning in small- to mid-scale large language models by automatically generating easier variants of reasoning problems and organizing them into difficulty buckets. It uses a self‑evolving bandit scheduler to adaptively allocate training, improving performance across math and multi‑hop reasoning tasks on 1–8 B models. LoT achieves significant gains (e.g., +32 pp on AddSub, +16 pp on QASC) and converges faster than staged curricula.
By Minghui Liu, Thomas Magelinski, Dehao Yuan, Qi Yu, Furong Huang
The paper introduces Network Feasibility Geometry Reinforcement Learning (NFG‑RL), a method that enforces multi‑layer network constraints—such as interference, power‑rate coupling, flow conservation, service chains, capacity, latency, and reliability—by transporting a proto‑policy through a differentiable feasibility map. By compiling heterogeneous constraints into typed residual blocks and using a variational transport operator, NFG‑RL ensures almost‑sure feasible execution and shapes exploration and gradients to respect active constraints. Experiments on two wireless‑edge surrogate environments show that NFG‑RL boosts feasible utility by 37.5–41.5 %, cuts raw‑action violations by 48.5–60.8 %, and reduces P99 delay by 57.0–75.5 % compared to leading baselines.
By Zuyuan Zhang, Zeyu Fang, Mahdi Imani, Nathaniel D. Bastian, Tian Lan
The paper introduces MGRL-RSCC, a multi‑granularity reward reinforcement learning framework for Remote Sensing Change Captioning (RSCC). It uses a CNN with hierarchical self‑attention to extract visual features, a Transformer decoder for visual‑to‑linguistic translation, and a dual‑decoding strategy combined with token‑level supervised learning and self‑critical reinforcement learning. Three reward functions—linguistic fluency, change state consistency, and structural‑semantic relevance—are employed to improve caption quality and reduce exposure bias and conservative generation.
By Futian Wang, Mengqi Wang, Xiao Wang, Wentao Wu, Haowen Wang, Zhicheng Zhao, Jin Tang
The paper introduces ELEMENT, a framework that combines episodic and lifelong entropy maximization to drive reward-free exploration in reinforcement learning. It addresses two key limitations of existing entropy-based methods: the vanishing intrinsic reward after a state is visited and the computational cost of estimating entropy over large datasets. ELEMENT achieves this by deriving an average episodic state entropy reward and employing a k‑NN graph‑based estimator for lifelong entropy, leading to superior state coverage and unsupervised pre‑training performance compared to current baselines.
By Hongming Li, Zhao Yang, Xiaoxuan Liang, Shujian Yu, Jose C. Principe
arXiv:2604. 00531v2 Announce Type: replace Abstract: Multi-task representation learning exploits the shared structure among related tasks by learning a common latent representation, thereby improving sample efficiency.
By Jiabin Lin, Shana Moothedath
TelecomGPT‑R1 is an open‑source family of unified telecom reasoning models that address the limitations of existing telecom‑specific and general‑purpose LLMs. It is built on a four‑axis framework—protocol, knowledge, modeling, and fault—and trained on a curated corpus of 104,880 verified question‑answer pairs with chain‑of‑thought reasoning. After supervised fine‑tuning, dynamic sampling policy optimization with task‑routed rubric rewards is used to stabilize reinforcement learning across heterogeneous telecom tasks, achieving an 89.64% mean score on seven GSMA Open Telco Leaderboard benchmarks, surpassing leading proprietary models.
By Bohao Wang, Chenwei Wu, Hang Zou, Yu Tian, Lina Bariah, Li Wei, Chongwen Huang, Yongliang Shen, Zhaoyang Zhang, Merouane Debbah
The paper introduces a four‑stage framework—SFT, PG‑CoT, Dynamic, and K‑RL—to improve large language models for Traditional Chinese Medicine prescription generation. It addresses three key gaps: lack of auditable reasoning (SR Gap), failure to adjust prescriptions over time (LA Gap), and non‑enforcement of absolute contraindication rules (SC Gap). Experiments on 12 fine‑tuned models and 6 zero‑shot baselines show that the framework, particularly a 7B Mistral model, outperforms zero‑shot GPT‑5 on all three TCM evaluation metrics.
By Zheng Chen, ZhiCheng Du, Haoxuan Li, Peiwu Qin
Informed Masking (IM) is a new technique for aligning Diffusion Large Language Models (dLLMs) with Reinforcement Learning (RL). It identifies a systematic upstream/downstream token structure in dLLM rollouts and shows that masking downstream tokens creates better subproblems for likelihood estimation. When integrated into three state‑of‑the‑art dLLM RL methods on LLaDA‑8B‑Instruct, IM yields up to 2.01%, 8.68%, and 5.77% relative average gains on math and planning benchmarks while improving training stability.
By Xiaoyi Yu, Enver Sangineto, Pei Fu, Fiorenzo Parascandolo, Wenhui Tan, Ruikang Zhang, Rita Cucchiara, Ruihua Song, Jian Luan
The paper proposes a new approach to large language model (LLM) reasoning that moves beyond naive repeated sampling. Instead of generating many independent solutions, it first samples problem‑specific concepts, hints, or strategies and conditions answer generation on them, producing a single trajectory of diverse concepts. A small concept generator is then trained via reinforcement learning to maximize downstream success, leading to significant improvements in pass@k on hard mathematical reasoning tasks compared to both naive sampling and concepts from larger untuned models, and the trained generator transfers to unseen answer generators, including those from different model families.
By Ismail Labiad, Matthieu Kowalski, Marc Schoenauer, R\'emi Munos, Julia Kempe
The paper introduces ConRad, a reinforcement learning framework that fine‑tunes large vision‑language models to generate calibrated verbalized confidence estimates for radiology reports. ConRad offers both a single report‑level confidence score and a sentence‑level variant, trained with the GRPO algorithm and logarithmic scoring rewards to encourage truthful self‑assessment. Experiments show significant calibration improvements over existing methods, and clinical evaluation indicates that report‑level scores align well with clinicians’ judgments, enabling targeted review of low‑confidence statements.
By David Bani-Harouni, Chantal Pellegrini, Julian L\"uers, Su Hwan Kim, Markus Baalmann, Benedikt Wiestler, Rickmer Braren, Nassir Navab, Matthias Keicher
The paper introduces DeToxR, a reinforcement‑learning‑enhanced large language model designed to support decision making in acute toxicology cases. It fuses unstructured narratives from paramedics and patients with structured vital‑sign data to predict co‑ingested substances across 14 classes. In preliminary validation, DeToxR outperforms baseline models, achieving higher micro‑F1 and recall scores for poison identification.
By Nico Oberl\"ander, David Bani-Harouni, Tobias Zellner, Nassir Navab, Florian Eyer, Matthias Keicher
The paper introduces a test‑time reinforcement learning framework for anomalous video understanding, addressing challenges such as unreliable pseudo‑labels, inadequate reward design, and collapsed group‑relative advantages. It proposes dual‑query consistency filtering, an entropy‑aware consensus reward, and a virtual negative anchor mechanism to improve sample reliability, reward quality, and policy‑gradient signals. Experiments on VAU‑Bench demonstrate significant performance gains, especially on the ECVA subset where accuracy rises from 75.81% to 90.00%.
By Huining Li, Yuxiang Duan, Jiyang Tan, Qian Li, MingCai Chen, Jian Zhang, Xingdong Sheng, Yuntao Du
GeoComposer is a new framework that improves photographic composition by generating textual guidance and visual exemplars for a given image. It uses a geometry-aware representation learning mechanism that incorporates geometric priors from a visual geometry foundation model to maintain global structure and fine-grained correspondences. A reinforcement learning strategy with a hybrid reward optimizes instruction following, aesthetic quality, and geometric consistency, leading to superior results compared to state‑of‑the‑art methods.
By Shuangzhi Li, Qiaoqiao Jia, Xingxin Chen, Guile Wu, Dongfeng Bai
arXiv:2609.22746v1 Announce Type: new
Abstract: Large Language Models (LLMs) have recently been introduced into traffic signal control (TSC) as decision agents due to their strengths in human-readabl...
By Huaitao Zhao, Tianlong Zhou, Weijie Wang, Jiasheng Shi, Weixiong Rao
arXiv:2609.23640v1 Announce Type: new
Abstract: Human-feedback alignment has made language models useful assistants and is commonly described as aligning them with humans. However, the responses peop...
By Suqin Yuan, Runqi Lin, Muyang Li, Guanzhe Hong, Jindong Gu, Lei Feng, Chris Russell, Tongliang Liu
arXiv:2609.23989v1 Announce Type: new
Abstract: Building general-purpose agents for industrial deployment requires integrating multiple capabilities, each typically acquired at a distinct stage of tr...
By Haixin Wang, Xiaoxuan Wang, Junkai Zhang, Han Zhang, Renliang Sun, Alexander K Taylor, Yidan Shi, Haoran Deng, Chenguang Wang, Jason Cong, Yizhou Sun, Wei Wang
arXiv:2609.24174v1 Announce Type: new
Abstract: Long-horizon language agents receive sparse terminal feedback, while intermediate rubrics provide structured but potentially misspecified assessments o...
By Xuchun Hu