AI agents improve task success by exploring multiple trajectories, but for computer-use agents each trajectory modifies external environment state. Branching from an intermediate point is correct only when restoration is observation-equivalent - future actions produce the same observations - and practical only when cre...
Jia-Kai Xu, Tian-Le Zhou, Georgios Liargkovas et al.· 0 citations
LLM agents can make unsafe tool calls even when instructed to behave safely. Existing defenses constrain agents before execution, modify tool inputs/outputs, or rely on LLM judges; these approaches may depend on model behavior or block unsafe actions without helping the agent recover. We argue that the execution enviro...
Charlie Summers, Prajwal Raghunath, Aaditya Pai et al.· 0 citations
This work studies parallel coding agents' failures with stale, a benchmark for semantic coordination, and finds that among 834 runs on 417 mined Django pairs, only one showed interference after correcting the grading procedure.
Hao-Chen Xia, Eugene Wu, Yongjoo Park· Proceedings of the 2nd Works...· 0 citations
FlowCheck, a constraint language to specify user-visible information flows directly through the application interface, where constraints can also be displayed and inspected without reading code, and are structured enough for reliable LLM generation.
Reya Vir, Lydia B. Chilton, Zhuo Zhang et al.· Proceedings of the 2nd ACM S...· 2 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
