paper-with-me

홈 › Papers

Constraint-Aware Reinforcement Learning via Adaptive Action Scaling

2025-10-13 · Murad Dawood, Usama Ahmed Siddiquie, Shahram Khorshidi, Maren Bennewitz arxiv

Safe reinforcement learning (RL) seeks to mitigate unsafe behaviors that arise from exploration during training by reducing constraint violations while maintaining task performance. Existing approaches typically rely on a single policy to jointly optimize reward and safety, which can cause instability due to conflicting objectives, or they use external safety filters that override actions and require prior system knowledge. In this paper, we propose a modular cost-aware regulator that scales the agent's actions based on predicted constraint violations, preserving exploration through smooth action modulation rather than overriding the policy. The regulator is trained to minimize constraint violations while avoiding degenerate suppression of actions. Our approach integrates seamlessly with off-policy RL methods such as SAC and TD3, and achieves state-of-the-art return-to-cost ratios on Safety Gym locomotion tasks with sparse costs, reducing constraint violations by up to 126 times while increasing returns by over an order of magnitude compared to prior methods.

📄 PDF Abstract BibTeX arXiv:2510.11491

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Adaptive Scaling of Policy Constraints for Offline Reinforcement Learning

2025-08-27 · Tan Jing, Xiaorui Li, Chao Yao, Xiaojuan Ban 외 arxiv

Offline reinforcement learning (RL) enables learning effective policies from fixed datasets without any environment interaction. Existing methods typically employ policy constraints to mitigate the distribution shift enc…

Reinforcement LearningOffline RL

3D Optimization for AI Inference Scaling: Balancing Accuracy, Cost, and Latency

2025-10-21 · Minseok Jung, Abhas Ricky, Muhammad Rameez Chatni arxiv

AI inference scaling is often tuned through 1D heuristics (a fixed reasoning pass) or 2D bivariate trade-offs (e.g., accuracy vs. compute), which fail to consider cost and latency constraints. We introduce a 3D optimizat…

Learning to Explore: Scaling Agentic Reasoning via Exploration-Aware Policy Optimization

2026-05-09 · Xingyuan Hua, Sheng Yue, Ju Ren arxiv

Recent advancements in agentic test-time scaling allow models to gather environmental feedback before committing to final actions. A key limitation of existing methods is that they typically employ undifferentiated explo…

Reinforcement Learning

QuantumQA: Enhancing Scientific Reasoning via Physics-Consistent Dataset and Verification-Aware Reinforcement Learning

2026-04-20 · Songxin Qu, Tai-Ping Sun, Yun-Jie Wang, Huan-Yu Liu 외 arxiv

Large language models (LLMs) show strong capabilities in general reasoning but typically lack reliability in scientific domains like quantum mechanics, which demand strict adherence to physical constraints. This limitati…

Reinforcement Learning

CORE: Constraint-Aware One-Step Reinforcement Learning for Simulation-Guided Neural Network Accelerator Design

2025-06-04 · Yifeng Xiao, Yurong Xu, Ning Yan, Masood Mortazavi 외

Simulation-based design space exploration (DSE) aims to efficiently optimize high-dimensional structured designs under complex constraints and expensive evaluation costs. Existing approaches, including heuristic and mult…

Reinforcement Learning (RL)