Large language model (LLM) agents solving multi-step tasks frequently commit to trajectories that are doomed to fail, yet continue to consume substantial inference compute before the failure becomes observable. We show that failure is predictable early from the agent's internal representations: lightweight per-round probes on hidden activations anticipate eventual episode failure as early as the first interaction round, where scorers reading only the agent's observable behavior are barely better than chance.
arXiv:2607. 06503v1 Announce Type: new Abstract: Large language model (LLM) agents solving multi-step tasks frequently commit to trajectories that are doomed to fail, yet continue to consume substantial inference compute before the failure becomes observable.
By Kai Ruan, Zihe Huang, Ziqi Zhou, Qianshan Wei, Xuan Wang, Hao Sun
The paper introduces Continual Search, an iterative framework that guides large language models to persistently search for diagnostic evidence in long AI agent execution logs, addressing the limitations of one-shot judgments. Evaluated on four existing RCA benchmarks and a new large-scale dataset called MegaRCA-Mix, Continual Search consistently boosts attribution performance, achieving a 40% F1 improvement for GPT‑5.5 on MegaRCA‑Mix. The results show that effective search can outweigh raw model scale, enabling lower-tier models to outperform higher-tier ones in root‑cause attribution tasks.
By Harsh Raj, David Lee, Anas Mahmoud, Renxiong Wang, Razvan-Gabriel Dumitru, Chenguang Wang, Tong Zhao, Yunzhong He, Darvin Yi, Vipul Gupta
arXiv:2606. 10794v3 Announce Type: replace Abstract: Existing black-box LLM provenance methods achieve comparability by querying every candidate model with the same diagnostic prompts.
By Jiaxu Liu, Sunnan Mu, Dong Huang, Liuyin Wang, Jing Shao, Jie Zhang
The paper discusses how the same computational task can require different reasoning strategies depending on the order in which evidence arrives, introducing the concept of an "information‑slack dilemma." It argues that early computation may be useful only if its benefits outweigh the costs of later verification, invalidation, and recovery, and proposes a research agenda focused on selective recovery and predictive policies. The authors emphasize evaluating these approaches by separating early‑execution effects, deployment value versus full‑input alternatives, and the added value of predictive policies while considering shared‑resource costs.
By Xiaotian Zhang (Trooly.AI)
Online A/B experiments are the decision standard for user engagement, but traffic and readout time limit how many conversational-AI changes can be tested. We ask whether an offline signal designed to...
arXiv:2608. 05235v1 Announce Type: cross Abstract: Research agents increasingly conduct multi-round machine-learning experiments in industrial recommendation settings and retain the resulting trajectories to guide later decisions.
By Zijie Zhuang, Changxin Lao, Pengbo Xu, Hanwen Xu, Ruochen Yang, Yingzhi He, Peng Zhang, Jiangxia Cao, Yusheng Huang, Guohong Mu, Jian Liang, Ruiming Tang, Shuang Yang, Zhaojie Liu, Wenwu Ou, Kun Gai
arXiv:2607. 17545v1 Announce Type: new Abstract: Language agents depend on memory across interactions.
By Qingcan Kang, Mingyang Liu, Shixiong Kai, Kaichao Liang, Zhentao Tang, Yuqi Cui, Tao Zhong, Mingxuan Yuan
The paper presents a layered framework for evaluating conversational AI by aligning offline proxy signals with online A/B experiment outcomes. It introduces a three‑step alignment chain—behavioral label to product outcome, classifier to candidate behavior, and offline signal to experiment effect—alongside an audit protocol that compares confidence intervals and rankings. In a real‑world deployment, the composite proxy achieved 81.1% F1 versus 34.3% for the raw classifier, correctly predicting direction on all 113 contrasts and enabling efficient prioritization of candidate models before costly online testing.
By Xuanyi Li, Vaskar Nath, Hossein Amirkhani, Jay Li, Alex Deng
The paper reports that language‑model agents used for customer‑relationship management can be misled by optimistic assertions from sales representatives in CRM records, leading to incorrect deal approvals. In a benchmark of 100 lead‑qualification tasks, models incorrectly cleared 29 of 31 deals where the representative’s claims contradicted company policies, with misalignment rates ranging from 87% to 97% across seven models. The authors propose diagnostic methods—including bucket analysis, same‑information controls, and compute‑step controls—to distinguish persuasion from information gaps and to quantify the impact of incentive‑misaligned witnesses.
By Rahul Balakavi
arXiv:2606. 05414v1 Announce Type: cross Abstract: Early failure alerting requires deciding, while a dialog or agent trajectory is still unfolding, whether to flag it as likely to fail.
By Avinash Baidya, Xinran Liang, Ruocheng Guo, Xiang Gao, Kamalika Das
arXiv:2607. 27083v1 Announce Type: new Abstract: As LLM agents increasingly depend on diverse external services such as search engines, databases, and connectors, agent harnesses face a fundamental tool-selection challenge: acquiring too few tools leaves the task under-informed, while too many adds cost, context load, and privacy exposure.
By Yicheng Feng, Yan Zhang, Yan Cheng, Wei Qi