The paper demonstrates that safety mechanisms for autonomous large language model agents fail to compose across iterative loops, as trajectory‑scoped monitors cannot detect attacks whose evidence is spread over multiple iterations. It introduces LoopHarness, a system that maintains a persistent, non‑decaying safety state across loops, bounding unauthorized actions with a constant that does not grow with the number of iterations. The authors provide a comprehensive evaluation protocol, including attacks that require cross‑iteration evidence, module ablations, and adaptive white‑box red‑team testing.
By Chenhao Wu, Haoxuan Jia, Yang Liu, Yingguang Yang, Yuhan Lin, Chongyang Zhang, Hao Zheng, Yulin Huang, Jianshen Zhang, Yongzhi Qi, Shang Luo, Kefu Xu, Jifeng Zhu, Bin Chong
arXiv:2609.06036v1 Announce Type: new
Abstract: Proposal-based controllers---learned policies, language-model planners, and other black-box \emph{generators}---are increasingly deployed behind runtim...
By Guangxi Wan, Yongbo Xie, Yuqi Liu, Qingwei Dong, Qingxin Li, Hongfei Bai, Peng Zeng
The paper introduces new evaluation metrics for safe reinforcement learning that go beyond average safety guarantees by examining how often and how severely safety bounds are violated, consistency across tasks and bounds, and the relationship between training-time and final policy behavior. It also proposes a safety tier system for categorizing algorithms and presents empirical safety evaluations on multiple navigation tasks. The authors recommend reporting aggregate metrics, distributional data, and task‑specific results together, and provide an open‑source suite, SafeRLEval, to facilitate reliable safety assessment.
By Lindsay Spoor, Aske Plaat, Thomas Moerland
arXiv:2606. 26057v1 Announce Type: cross Abstract: AI agents are granted access to tools, APIs, and other infrastructure, making them active principals in those systems.
By Seth Dobrin, {\L}ukasz Chmiel
The paper introduces neuro‑symbolic computer use, a method that learns reusable policies to execute recurring computer workflows efficiently. Instead of re‑planning each run, the learned policy encodes stable decisions (ordering, variables, loops, branches) into executable code while delegating observation‑dependent decisions to neural models. Using neuro‑symbolic policy iteration, the approach iteratively refines the policy from a single agent trajectory, diagnoses failures, and revises the code with a coding model, achieving superior Pass^3 scores and significant reductions in per‑run cost and latency on OSWorld‑Verified and ScienceBoard benchmarks.
By Hyewon Suh, Thanh Minh Nguyen, Chih-Lun Lee, Darrow Hartman, Lizhao Liu, Xin Eric Wang, Ang Li, Jiachen Yang
arXiv:2607. 15003v1 Announce Type: new Abstract: The deployment of autonomous cyber-physical systems in safety-critical environments requires closed-loop control strategies (i.
By Riccardo Curcio, Toni Mancini, Enrico Tronci
arXiv:2606. 04634v1 Announce Type: new Abstract: Trust in a decision-making system requires both safety guarantees and the ability to interpret and understand its behavior.
By Sabine Rieder, Stefan Pranger, Debraj Chakraborty, Jan K\v{r}et\'insk\'y, Bettina K\"onighofer
arXiv:2606. 31320v1 Announce Type: new Abstract: Safe online reinforcement learning requires policies to respect safety constraints while maintaining smooth optimization dynamics.
By Hongpeng Cao, Liqun Zhao, Yuliang Gu, Naira Hovakimyan, Lui Sha, Marco Caccamo
arXiv:2607. 22868v1 Announce Type: new Abstract: Runtime guardrails act before irreversible tool calls, but their guarantees depend on what policy state is representable, what a judge observes, and whether intervention changes future behavior.
By Shawn Ray
arXiv:2606. 00270v1 Announce Type: new Abstract: Shielding is an effective approach to formally guarantee the safety of reinforcement learning agents in Markov decision processes (MDPs).
By Edwin Hamel-De le Court, Thom Badings, Alessandro Abate, Francesco Belardinelli, Francesco Fabiano
arXiv:2608. 03662v1 Announce Type: new Abstract: Safety shields are runtime enforcement mechanisms that restrict the actions of a controller to guarantee safety.
By Filip Cano, Thomas A. Henzinger, Konstantin Kueffner
RePolicy is a reinforcement learning approach designed to invoke safety policies for language model agents by evaluating entire execution trajectories within context-dependent policy libraries. It generates policy-grounded rationales and safety judgments, and is initialized with the PolicyTraj-20K dataset before fine-tuning via GRPO with verifiable rewards and policy-context perturbation. Experiments on six safety benchmarks demonstrate strong safety-detection performance and robust policy invocation across varying contexts.
By Houcheng Jiang, Boxuan Zhang, Qiyong Zhong, Junfeng Fang, Xiang Wang, Xiangnan He