The paper introduces a new optimization method called Continuous‑Delayed‑Memory Stochastic Gradient Descent, which incorporates past states of the iteration process to improve exploration and convergence compared to vanilla SGD. It also proposes a continuous‑time reinforcement learning framework that uses policy gradients for exploratory policies, showing that its optimality conditions recover the Gibbs policy from earlier studies. The work builds on stochastic differential equations adapted with neural network parameterizations to model quasars’ stochastic brightness variations from light‑curve data.
By Debartha Paul, Juncheng Yi
The paper introduces PARTS, a real‑world subtask reinforcement learning framework that fine‑tunes a pretrained robot policy by focusing on critical bottleneck subtasks while keeping the base policy frozen. It uses agent‑generated selectors and success verifiers to provide local rewards, enabling learning even when full‑task successes are rare. Experiments on bimanual YAM and single‑arm Franka robots show that PARTS raises complete‑task success from 32% to 61% and from 50% to 95%, respectively, with only tens of minutes of real‑world RL rollouts and minimal human intervention.
By Sichang Su, Benjamin Yang, Zhiyun Deng, Boyuan Liang, Yip Fun Yeung, Zelin Wang, Lingfeng Sun
arXiv:2607.08837v4 Announce Type: replace-cross
Abstract: Exploration is essential to RL since a policy cannot improve by repeatedly sampling the behaviors it already prefers. Standard methods inject...
By Sunshine Jiang, John Marangola, David Zhang, Raghuram Kowdeed, Ruiyang Luo, Nitish Dashora, Richard Li, Pulkit Agrawal, Zhang-Wei Hong
arXiv:2609.21108v1 Announce Type: new
Abstract: Deep reinforcement learning (DRL) has achieved strong performance across a wide range of continuous-control problems. These continuous-control policies...
By Sachini Weerasekara, Sagar Kamarthi, Jacqueline Isaacs
arXiv:2609.21945v1 Announce Type: new
Abstract: Urban transportation networks present complex optimization challenges spanning calibration of high-fidelity simulators and real-time operational contro...
By Adewumi Augustine Adepitan, Christopher J. Haruna, Oluwasegun Adegoke, Ayooluwatomiwa Ajiboye, Oluwatobi Oluwasakin
arXiv:2609.21976v1 Announce Type: cross
Abstract: We introduce Multiplicatively Optimistic Regret Matching (MORM), an uncoupled learning rule for finite general-sum games. Under simultaneous full-inf...
By Ashkan Soleymani, Georgios Piliouras
arXiv:2505.04193v2 Announce Type: replace
Abstract: Simplicity is a critical inductive bias for designing data-driven controllers, especially when robustness is important. Despite the impressive resu...
By Bang You, Chenxu Wang, Wenju Yang, Di Guo, Huaping Liu
arXiv:2604.00200v2 Announce Type: replace
Abstract: We study offline constrained reinforcement learning from human feedback with multiple preference oracles. Motivated by applications that trade off...
By Brenden Latham, Mehrdad Moharrami
arXiv:2609.21675v1 Announce Type: new
Abstract: Despite the remarkable progress in Multimodal Large Language Models (MLLMs), prevailing Chain-of-Thought (CoT) paradigms remain confined to the natural...
By Wan Xu, Yuanfan Guo, Kevin Han, LaLa Chen, Wangmeng Zuo
arXiv:2609.22083v1 Announce Type: new
Abstract: We present MintAct, a family of vision-language models that unifies UI grounding, multi-step navigation across mobile, desktop, and web, and visual too...
By Mingfei Gao, Rui Tian, Haiming Gang, Bohan Zhai, Le Zhang, Yuanzheng Gong, Di Feng, Ege \"Ozsoy, Kaixin Ma, Vishwesh Kirthivasan, O\u{g}uzhan Fatih Kar, Roman Bachmann, Anders Boesen Lindbo Larsen, Afshin Dehghan
arXiv:2609.20842v1 Announce Type: new
Abstract: Text-to-SQL translates natural-language questions into executable SQL queries, but open-source large language models still require task-specific post-t...
By Qifeng Cai, Xuanguang Pan, Hao Liang, Chang Xu, Wentao Zhang
arXiv:2609.20844v1 Announce Type: new
Abstract: Deepresearch (DR) agents interact with real-world web environments through multi-turn search and visit, causing their contexts to grow rapidly over tim...
By Zihan Wang, Hao Wang, Boyuan Jiang, Yiqun Zhang, Shi Feng, Xiaocui Yang, Yiwen Ye, Jianghang Lin, Xiaozhong Ji, Jinghao Lin, Kai Wu
arXiv:2609.21231v1 Announce Type: new
Abstract: Reference-based metrics for Grammatical Error Correction (GEC) such as M$^2$ and ERRANT assume that the reference set enumerates all valid edits, and t...
By Ruotian Wu, Bill E. Johnson, Gene Saunders, Osama Hamzeh, Ankit Vadehra, Pascal Poupart
arXiv:2609.20886v1 Announce Type: cross
Abstract: Business intelligence (BI) is a cornerstone of enterprise decision-making and is widely used by enterprise users in software such as Power BI and Tab...
By Chuxuan Hu, Yeye He, Penny Zhou, Wee Hyong Tok, Daniel Kang, Surajit Chaudhuri
Reinforcement Learning with Verifiable Rewards (RLVR) is expanding from tasks with well-defined correctness signals, such as mathematics and code, toward multifaceted quality requirements specified by...
Reinforcement learning (RL) is vital for optimizing video generation models, with a robust reward model (RM) serving as the cornerstone. However, existing video reward models often produce unstable sc...
Model-free reinforcement learning can acquire contact-rich robotic manipulation skills through trial-and-error interaction, but it often requires the policy to learn both task strategy and low-level m...
UniPolicy is a unified objective‑specific policy framework for search advertising that jointly optimizes relevance, click propensity, and commercial value. It uses objective‑aware prefix tokens, sparse MoE‑LoRA routing, and residual FFNs to decouple parameters within a shared backbone, and constructs pairwise preferences from multi‑stage behavioral feedback to strengthen clicked candidates. In large‑scale offline tests and a 7‑day online A/B test, UniPolicy improves CTR by 0.71%, RPS by 1.58%, and advertising revenue by 1.32% while keeping serving latency stable.
By Kun Yao, Yuhang Zhou, Yichi Zhang, Zeliang Tong, Shengri Xue, Haitao Wang, Siyu Lu, Qianlong Xie, Xingxing Wang
The paper introduces a dependency‑graph framework to formalize compositional reasoning in language models, defining three increasing levels of compositionality. Using data‑structure tasks with deterministic rewards, the authors observe a consistent asymmetry: training on decomposed skills does not reliably transfer to composed tasks, whereas training on composed tasks transfers more readily to decomposed ones. They provide a theoretical explanation for this asymmetry and evaluate its effects under length extrapolation, structural distribution shift, and transfer to unseen skills, concluding with a pilot study on real‑world tool‑calling benchmarks that suggests the phenomenon extends to practical settings.
By Yu He, Yingxi Li, Yifei Wang, Ellen Vitercik
The paper investigates how a reference policy can be used to steer regularized self‑play toward a specific equilibrium in two‑player zero‑sum games. By anchoring the reference at a target equilibrium and refining the self‑play process, the authors achieve precise convergence to that target with very low exploitability and coordinate error. The study also explores the effects of off‑manifold references, mirror‑step sizing, and boundary saturation on selection accuracy.
By Luis Leal