Modern LLM agents persist state across sessions via two channels: long-lived memory files silently prepended to the system prompt, and scheduled tasks waking the agent without user input. A one-time indirect prompt injection committed to either channel can re-fire across sessions long after the attacker leaves, threats...
Zhi-Yu Xu, Xu-Kai Wang, Zhiyuan Huang et al.· 0 citations
This work proposes APTER (Adaptive Post-Training with Expert-Grounded Rubrics), a framework that integrates structured domain knowledge into fine-grained evaluation, optimization, and diagnosis for specialized complex reasoning and shows consistent gains across both domains.
Xu-Kai Wang, Liangqi Li, Zhi-Yu Xu et al.· 0 citations
This work identifies Teacher Uncertainty Contraction (TUC), a systematic phenomenon whereby the teacher's predictive uncertainty decreases as it continues from a student-generated prefix, and proposes Adaptive-Continuations On-Policy Distillation (AC-OPD), which augments informative states along student rollouts with t...
Jin-Gang Zhou, Yu-Yi Zhou, Hai-Yang Guo et al.· 3 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.