The paper investigates how language‑model judges can make version‑dependent errors when evaluating upgraded agents. Using 35 public coding‑agent submissions, two customer‑service agents, and over a thousand expert‑labeled trajectories, the authors show that fixed judges often reject task‑conditioned error invariance and can incorrectly approve failed patches, especially as agent capability increases. Paired audits of current outputs reduce interval width only marginally, and the study concludes that independent human patch review is still necessary.
By Jiapeng Li
arXiv:2606. 10241v1 Announce Type: new Abstract: Autonomous improvement loops are hard to trust because the improvement process is usually external scaffolding bolted onto the agent: failures go unlogged, diagnoses cannot be replayed, and promote-or-discard decisions land in a side database rather than the agent's own history.
By Yohei Nakajima
The paper introduces a framework for evaluating AI systems that not only checks final labels but also tracks the reasoning behind them through three core sources—grounds, norms, and authority—forming an eight-cell counterfactual judgment cube. It defines minimal source replacement sets, called judgment receipts, to explain changes in verdicts and provides certification cost bounds for black-box evaluators. The authors present ReasonBench, a benchmark with 19,520 cases, and demonstrate that while high standard accuracy can mask robustness issues, receipt accuracy reveals significant gaps in reasoning consistency across different models.
By Ye Chen, Weining Zhang
The Agent Error Dataset (AED) presents 50,228 error–diagnosis pairs collected from 9,961 source tasks across 33 environments, 19 harness families, and 23 policy models in text‑based agent systems. A five‑stage Agentic Error‑to‑Training (AET) pipeline generates diagnoses and proposed corrections, verifies them against recorded evidence, and creates separate training views for diagnosis and actor recovery. Experiments show that first‑proposal corrections improve verifier pass rates from 18.4% to 51.1%, and fine‑tuning with full‑diagnosis data raises Qwen3‑8B’s exact‑step agreement from 47.2% to 63.6% on a holdout set.
By Kunlun Zhu, Xuyan Ye, Yibo Li, Cheng Qian, Beibin Li, Heng Ji
arXiv:2607. 28871v1 Announce Type: cross Abstract: When a repair agent runs a test and sees it pass, the result is treated as evidence about the reported defect.
By Xiaonan Xu, Wenjing Wu
FinalityBench is an executable benchmark that tests how agents decide on shipping, re‑capturing, refunding, or waiting when a merchant’s payment processor, ledger, ERP, and bank feed receive delayed, duplicated, dropped, or reordered messages, causing contradictory beliefs about an order. The benchmark uses a hidden canonical event log and faulted delivery streams to generate system views, scoring each episode by the merchant’s terminal economic position relative to a privileged reference. It contains 321 tasks, including 45 twin pairs where all four views are identical yet the correct disposition differs, and evaluates nine programmatic policies, revealing that a ship‑on‑first‑sign policy performs best by accuracy but worst by paired loss, while a runtime‑gated irreversible‑action policy achieves 85.4% accuracy without losing money.
By Abhishek Sharma
arXiv:2609.37315v1 Announce Type: cross
Abstract: Tool-using agents are entering settings where a wrong action carries real cost, and the benchmarks certifying them grade what each simulated tool cal...
By Rohith Reddy Bellibatlu, Zichong Wang, Wenbin Zhang
The paper "trajectory-judge: What Outcome-Only LLM Judges Miss on Agent Trajectories" examines the limitations of outcome-only evaluation for large language model agents. Using a deterministic tool‑using support‑desk environment with a scripted oracle policy and a fault injector, the authors compare five different judging approaches—programmatic rules, outcome‑only, step‑rubric at two model sizes, and a self‑consistency ensemble—on metrics such as detection, step localisation, fault typing, calibration, and cost across 400 trajectories. The study finds that outcome‑only judges miss many silent faults and generate false positives, while step‑rubric judges achieve higher recall with no false alarms but at greater cost, and that none of the judges read the final reply, allowing fabricated promises to evade detection.
"whyItMatters":"The findings highlight that current production‑default outcome‑only evaluations can overlook critical failures in agent behavior, underscoring the need for more nuanced, step‑level judging methods to ensure reliable LLM agent performance."
By Hadi Mohammadi
arXiv:2609.24194v1 Announce Type: new
Abstract: Evaluation scores used around LLM systems -- including reward models, rerankers, and LLM judges -- can track surface form instead of the quality they c...
By Daein Weon, Dongho Kang
ContractRL introduces a contract-constrained sequential repair protocol for structured tool calls, modeling verifier-guided JSON repair as a bounded decision process. The policy observes candidate data, verifier feedback, JSON pointers, repair history, and budget, using a contract-derived action mask to filter invalid operations before a deterministic validator applies changes. Compared to Patch‑SFT and full regeneration, ContractRL achieves higher semantic success (0.9362 vs. 0.9076 and 0.9148) while generating fewer tokens (34.4 vs. 44.9 and 137.2), and policy optimization further improves success rates.
By Miaobo Hu, Shuhao Hu, Xiaobo Guo, Xin Wang, Bokun Wang, Yina Sa, Daren Zha, Jun Xiao
arXiv:2608. 05212v1 Announce Type: new Abstract: Deep search agents tackle challenging questions through long-horizon web interactions, a process that is both complex and fragile: small reasoning errors may propagate through long, noisy trajectories into fluent but incorrect answers.
By Zhixiang Liang, Yifei Liu, Yidan Huang, Haozhe Zhao, Beichen Huang, Jiaqi Wang, Nan Duan, Qiong Cao
arXiv:2607. 19453v1 Announce Type: cross Abstract: We audit whether candle-based machine-learning models can turn predictions of cryptocurrency extrema or short-horizon outcomes into positive Binance Spot paper policies after assumed costs.
By Ayoub Jadouli