Repair or Resample? Rethinking Failure Debugging in LLM Multi-Agent Systems
Score
76首次系统评估重跑式修复是否真正因果有效，并提出了可控复现框架和新数据集，填补了该领域空白。
研究设计严谨，包含大规模人工标注数据和多框架实证，论证充分且数据支撑有力。
为多智能体系统开发者提供了可用的故障复现与修复工具，能直接指导调试实践。
针对当前 Agent 应用爆发但可靠性不足的痛点，研究成果高度贴合当下需求。
来自 arXiv 学术预印本，虽未经同行评审，但方法规范、数据详实，可信度较高。
Base 84.5 × T2 Source weight ×0.9 = 76
Curation line 70Selected
Why selected
The paper introduces a new framework, a new dataset, and systematic empirical comparisons, providing clear evidence and significant incremental value for agent reliability research.
AI commentary
「This paper debunks an industry illusion: many teams fix agent failures simply by retrying. The numbers are sobering—rerun repair rate is only 6.9%, while symptom-driven targeted intervention nearly triples it. For those building multi-agent products, SymTrace and SymFail offer practical blueprints for building regression sets without relying on luck.」
Quick overview
As large language model-based multi-agent systems (MASs) are increasingly applied to long-horizon complex tasks, their reliability has become the core bottleneck hindering real-world deployment. Existing debugging and repair methods typically rely on rerunning the entire execution trajectory and resampling, but whether these methods cause causal repairs or merely stochastic repairs by leveraging LLM sampling randomness remains an open question. To evaluate the effectiveness of MAS repair methods, this paper introduces SymTrace, a controlled evaluation framework that records MAS execution traces and establishes intervention anchors. During replay, it reconstructs the pre-anchor execution from logs and only regenerates the downstream trajectory, enabling reliable reproduction of MAS failures. The authors also construct SymFail, a dataset of 536 human-annotated failure trajectories with graph-linked locations, categories, and trace evidence. A large-scale empirical study across three mainstream MAS frameworks reveals that existing unguided rerun methods are highly unreliable, with low failure reproduction and repair rates of only 67.97% and 6.90%, respectively. Building on these findings, they further explore a symptom-driven intervention method that successfully repairs 20.15% of failed cases, a 191.89% improvement over state-of-the-art repair methods. This study provides actionable insights for MAS debugging and repair research, paving the way for robust deployment of multi-agent systems.
- SymTrace provides a controlled evaluation framework that records execution traces and uses intervention anchors to reliably reproduce MAS failures via log replay and downstream regeneration.
- SymFail dataset includes 536 human-annotated failure trajectories with graph-linked locations, categories, and trace evidence.
- Empirical results show unguided rerun methods are unreliable: 67.97% failure reproduction and 6.90% repair rate; symptom-driven intervention raises repair to 20.15%, a 191.89% improvement over SOTA.
Full text
This is an AI-generated Chinese digest; use "View original" for the source text.
Related items
85
FeaturedSetoka: A Benchmark for Hierarchical User Understanding in Personalized Agents over Heterogeneous Data
Setoka：面向异构数据个性化Agent的层级用户理解基准
This paper is well worth attention, especially for researchers and engineers working on personalized agents, dialogue systems, and memory-augmented AI. It not only provides a benchmark but defines levels of user understanding from psychology, revealing systematic failures of current systems at higher levels. If you are designing memory-based agents, this paper reminds you that memory systems need to support cross-temporal, cross-source abstract reasoning, not just retrieval.
85
FeaturedBAND: Bayesian Network Distribution Regression for High-Dimensional Distribution Estimation – Breaking the Curse
BAND：高维分布估计的贝叶斯网络回归方法——打破维度诅咒
This paper is a must-read for anyone working on high-dimensional statistics, probabilistic modeling, and generative AI. It provides a simple yet powerful theoretical answer: sparsity can break the curse of dimensionality, and Bayesian networks offer a natural framework. Compared to black-box methods like VAEs and GANs, BAND is interpretable and provably convergent, especially valuable for industry needs like reliable confidence intervals. If you research time series forecasting or risk management, this paper may change how you approach high-dimensional density estimation.
84
FeaturedCycle-Consistent and Uncertainty-Aware Neural Surrogates for Tokamak Edge Plasmas
循环一致与不确定性感知的托卡马克边缘等离子体神经代理模型
This paper is worth attention because it demonstrates how advanced AI techniques (cycle consistency, uncertainty quantification) directly address a core computational bottleneck in nuclear fusion. For fusion researchers and engineers, it offers a practical tool to speed up simulations by orders of magnitude without sacrificing accuracy; for AI practitioners, it provides an excellent case of self-supervised learning and inverse problem solving in scientific applications.
83
FeaturedEquivariant Learning of a Transferable Three-Dimensional Classical Density Functional
等变学习可迁移的三维经典密度泛函
The most striking aspect of this paper is its direct coupling of machine learning to a core problem in statistical mechanics, using equivariant networks to learn a transferable density functional without any labels. For researchers in liquid physics or soft matter, this could be a new paradigm: do we need to run separate simulations for every new system anymore? For AI researchers, it also shows how to embed physical symmetries and variational principles into models rather than just fitting data. I recommend focusing on the methods, especially how equivariance and variational consistency are implemented—that's the real essence.
83
FeaturedComputational and Statistical Guarantees of the c-Rectified Flow
c-Rectified Flow 的计算与统计保证
This paper moves rectified flow from 'works in practice' to 'theoretically controllable,' which is especially important for those working on the underlying algorithms of generative models. If you are tuning pipelines like FLUX or SD3, understanding that ordinary rectified flow can fail when covariances do not commute may help you avoid hidden training pitfalls.
83
FeaturedLearning Deployable Manipulation Policies from High-Fidelity UMI Data Alone
从高保真UMI数据学习可部署操作策略
The most striking point of this paper is proving that pure robot-free data can match real teleoperation performance, breaking the convention of needing a small real-robot anchor for post-training. For research groups looking to reduce hardware costs and industrial robot application developers, HiFi-UMI's data production system and open-source dataset are directly usable resources. Pay attention to the engineering details of head-mounted SLAM and microsecond synchronization—they are key to achieving the reported accuracy.
83
FeaturedCapital Markets LLM Reliability Score: From Plausible to Bankable
资本市场LLM可靠性评分：从“看似合理”到“可银行化”
This paper is valuable for both finance and AI communities: it provides an actionable, workflow-oriented evaluation framework that's closer to real needs than generic benchmarks. If you're deploying LLMs in fintech or responsible for explainability/auditability, this is a must-read. The finding on decision usefulness is particularly striking—judges agree it's critical, yet models vary the most on it.
83
FeaturedWhere Should Optimizer State Live? Tiered State Allocation for Memory-Efficient Mixture-of-Experts Training
优化器状态应放在哪里？面向内存高效MoE训练的分层状态分配
This paper is highly valuable for teams training MoE models: it demonstrates that with careful optimizer state allocation, one can dramatically reduce memory without sacrificing accuracy—and even improve performance. It is especially relevant for labs or enterprises constrained by 40 GB or 80 GB GPU memory. The tiered strategy and ablation on momentum retention are worth studying for optimizing training frameworks like DeepSpeed or Megatron.