ATLAS is a dual‑horizon diagnostic evaluation framework designed for industrial tool‑use agents on web‑scale service platforms. It combines within‑turn trajectory‑wise signals (thinking, reflection, tool execution, response generation) with across‑turn user‑wise signals over multiple lifecycles to pinpoint execution failures and unmet capabilities. The framework employs rule‑based checks and calibrated LLM judges, distilling them into compact diagnostic models that enable developers to attribute failures, prioritize fixes, and reassess improvements. In Meituan Xiaotuan, ATLAS‑guided agents achieved higher recommendation relevance, factual reliability, user engagement (+2.84% answer read‑through rate), and order volume (+6.13% paid orders), leading to full production deployment.
By Wei Chen, Peilun Zhou, Zhaoyu Hu, Jiajun Chai, Derong Xu, Zhongni Hou, Yufei Zhang, Guojun Yin, Wei Lin, Zhi Zheng, Tong Xu
AuditKIT is a source‑available evaluation library that unifies benchmark accuracy, judge‑based assessment, runtime performance, tool‑use behavior, and retrieval quality into a single result representation. It offers direction‑aware metric contracts, configuration fingerprints, and extensible metrics and backends, while providing turn‑aware tool‑call matching and distinguishing verified task outcomes from trajectory‑level diagnostics. Seven case studies demonstrate how answer extraction, runtime behavior, and metric selection can alter evaluation outcomes across model comparison, serving runtimes, retrieval, tool calling, and interoperability.
By Shreeyans Arora, Utsav Avaiya, Zera Lyngkhoi, Hem Gosalia, Ram Mohan Rao Kadiyala, Vinay Kumar Sankarapu, Pratinav Seth
The paper introduces a framework that links the foundational capabilities of large language models (LLMs) to their agentic performance by measuring operation-level gaps. It constructs foundational profiles from benchmark data and operation profiles from real interaction trajectories, then quantifies how much observed operation performance deviates from predictions based on foundational capabilities. Experiments on fifteen model services in the τ²-Bench demonstrate that foundational profiles alone only partially predict agentic outcomes, while incorporating operation profiles significantly improves task-score predictions and reveals consistent execution gaps across task subsets.
By Wei Guo, Yuanjian Xu, Jianing Hao, Yongyao Duan, Ran Chen, Zhong Li, Guang Zhang
UniACE is a unified framework that standardizes the evaluation of large language model (LLM) agents by representing each benchmark as an instruction–tool–environment triplet and running models through a shared, task‑agnostic harness in isolated runtimes. It preserves native success criteria, offers an offline mode for dynamic‑resource tasks, and standardizes efficiency metrics, execution records, and failure attribution. Applying UniACE to 7 benchmarks across 24 domains and 15 models revealed significant score shifts, ranking reversals, and sensitivity to evidence representation, highlighting the impact of evaluation configuration on reported agent performance.
By Pengyu Zhu, Lijun Li, Yaxing Lyu, Qianxin Luo, Jingyi Yang, Yi Liu, Tingfeng Hui, Xinyu Yuan, Li Sun, Sen Su, Jing Shao
ToolRobustBench is a stage-wise diagnostic benchmark designed to evaluate and diagnose failures in tool‑calling agents, which are large language models that select tools, provide structured arguments, and interpret tool feedback. The benchmark aligns four perturbation families—tool‑interface, user‑intent, tool‑output/observation, and runtime‑environment—with the tool‑use pipeline, attributing failures to specific stages such as tool selection, schema grounding, argument binding, and feedback handling. Experiments across 15,456 instances, 7 models, and 16 local tools reveal that while overall performance is high, robustness degrades significantly, especially under tool‑output/observation perturbations, and mixed‑family perturbations produce non‑additive failure patterns.
By YiShan Zheng, Yuan Wu, Yi Chang
arXiv:2603. 14465v2 Announce Type: replace Abstract: While Large Language Models (LLMs) have evolved into tool-using agents, they remain brittle in long-horizon interactions.
By Shengda Fan, Xuyan Ye, Yupeng Huo, Zhi-Yuan Chen, Yiju Guo, Shenzhi Yang, Wenkai Yang, Shuqi Ye, Jingwen Chen, Haotian Chen, Xin Cong, Yankai Lin
CivBench is an open‑source benchmark that evaluates language‑model agents in the long‑horizon, tool‑mediated game Civilization VI using the Model Context Protocol (MCP). Each episode lasts over 300 turns, generating thousands of tool calls across a 76‑tool action space, and includes a narration layer that translates visual game state into structured text. The study characterises agent behaviour across four model families, introducing Proactive Monitoring Rate (PMR) and RAG@10 as interface‑level metrics, and finds that agents often under‑monitor strategic state and fail to execute near‑term commitments despite tool access and explicit guidance.
By Austin Tudor David Andrews, Liam Wilkinson, Jamie Heagerty, Harry Coppock, Jakob Nicolaus Foerster, Rui Ponte Costa
arXiv:2610.02330v1 Announce Type: new
Abstract: Large language models (LLMs) rely on long-horizon tool invocation sequences for complex tasks, where each invocation can alter the task state and condi...
By Yu Li, Zheng Zhang, Xin Liu, Shengtian Yang, Guangfeng Cai, Lei Feng
arXiv:2606. 01365v1 Announce Type: new Abstract: Tool-using multi-agent large language model (LLM) systems spend computation through model tokens, tool calls, retries, and code execution before producing an answer.
By Xianyou Li, Weiran Yan, Yichao Wu, Penghao Liang, Mengwei Yuan, Jianan Liu, Jing Yang
arXiv:2606. 15508v1 Announce Type: new Abstract: Tool-augmented large language model agents increasingly operate over large tool libraries, but existing evaluations often focus on whether a model can call a tool correctly rather than how the visible tool menu shapes reliability, efficiency, and safety-relevant risk exposure.
By Rahul Suresh Babu, Laxmipriya Ganesh Iyer
The paper investigates why large language model (LLM) agents fail on long, multi‑step production workflows despite high benchmark success. By testing nine models (1.2 B–671 B parameters) across six task families and multiple horizons, the authors find that task success follows a geometric decay governed by a per‑step reliability that never reaches 1, leading to inevitable collapse for long horizons. The degradation is driven mainly by step count rather than context length, and the study quantifies a significant gap between benchmark and production performance, especially for agentic tool‑use tasks.
By Shubhra Mittal
arXiv:2610.16333v1 Announce Type: new
Abstract: As AI agents increasingly act through external tools, tool results are no longer simple outputs but execution artifacts that must remain correctly asso...
By Yutong Zhi, Jie Zhang, Xiuqi Lu, Jiayuan Xu, Jing Shao