# Seeing the Whole Elephant: A Benchmark for Failure Attribution in LLM-based Multi-Agent Systems

arXiv: https://arxiv.org/abs/2604.22708 · HTML: https://mlsanity.com/paper/2604.22708/

## Key facts

- Cited by 22 papers in the map (data through 7 Oct 2026).
- Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use.
- Builds on 8 of its 25 references that are in the map.
- First posted on arXiv 24 Apr 2026.

## Placed in

- LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/)

## Builds on (most cited first, 8 of 8)

- [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://mlsanity.com/paper/2405.15793/) (2024)
- [CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Society](https://mlsanity.com/paper/2303.17760/) (2023)
- [Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks](https://mlsanity.com/paper/2411.04468/) (2024)
- [A Survey on LLM-based Multi-Agent System: Recent Advances and New Frontiers in Application](https://mlsanity.com/paper/2412.17481/) (2024)
- [Who is Introducing the Failure? Automatically Attributing Failures of Multi-Agent Systems via Spectrum Analysis](https://mlsanity.com/paper/2509.13782/) (2025)
- [Where Did It All Go Wrong? A Hierarchical Look into Multi-Agent Error Attribution](https://mlsanity.com/paper/2510.04886/) (2025)
- [GraphTracer: Graph-Guided Failure Tracing in LLM Agents for Robust Multi-Turn Deep Search](https://mlsanity.com/paper/2510.10581/) (2025)
- [Evolution in Simulation: AI-Agent School with Dual Memory for High-Fidelity Educational Dynamics](https://mlsanity.com/paper/2510.11290/) (2025)

## Built on by (most cited first, 22 of 22)

- [Where Do Deep-Research Agents Go Wrong? Span-Level Error Localization in Agent Trajectories](https://mlsanity.com/paper/2606.02060/) (2026)
- [Beyond Individual Intelligence: Surveying Collaboration, Failure Attribution, and Self-Evolution in LLM-based Multi-Agent Systems](https://mlsanity.com/paper/2605.14892/) (2026)
- [TRAJDEBUG: Tracing Error Lifecycle to Identify Critical Failures in Long-Horizon Agent Trajectories](https://mlsanity.com/paper/2608.06346/) (2026)
- [Long-Horizon Agent Trajectory Attribution: A Unified Benchmark and Fine-Grained Annotation Framework](https://mlsanity.com/paper/2608.06909/) (2026)
- [DiagEval: Trajectory-Conditioned Diagnosis for Reliable Software Evaluation with GUI Agents](https://mlsanity.com/paper/2605.17439/) (2026)
- [Evolve as a Team: Collaborative Self-Evolution for LLM-based Multi-Agent Systems](https://mlsanity.com/paper/2605.29790/) (2026)
- [Knowledge-Based Zero-Replay Debugging of Multi-Agent LLM Traces](https://mlsanity.com/paper/2606.14805/) (2026)
- [CatchBench: When Can an Agent Failure Be Caught?](https://mlsanity.com/paper/2608.22808/) (2026)
- [TelemetrySuffBench: Is Agent Telemetry Sufficient for Failure-Origin Diagnosis?](https://mlsanity.com/paper/2608.07899/) (2026)
- [Evidence-Calibrated Runtime Reconstruction for Agent Skills Across Heterogeneous Coding Agents](https://mlsanity.com/paper/2608.08793/) (2026)
- [ASCon: A Direction-Aware Reciprocal Agent--Step Contextualization Model for Failure Attribution in Multi-Agent Systems](https://mlsanity.com/paper/2608.10646/) (2026)
- [LongRCA Bench: Root-Cause Localization in Long-Horizon Agent Trajectories](https://mlsanity.com/paper/2608.15242/) (2026)
- [AgentGrad: Intervention-guided Prompt Optimization for Multi Agent Systems](https://mlsanity.com/paper/2609.08572/) (2026)
- [Diagnosis-Driven Automatic Repair for Agentic Workflow via Symbolic Inference](https://mlsanity.com/paper/2607.02882/) (2026)
- [What Process Evaluation of Coding Agents Actually Measures: Action, Task, and Step Are Three Different Levels](https://mlsanity.com/paper/2608.22960/) (2026)
- [Localizing Emergent Failures in Agentic AI: Recovering Minimal Repair Families via Counterfactual Replay](https://mlsanity.com/paper/2608.29228/) (2026)
- [Hearsay: Can an Auditor Trust the Record a Deployed Agent Harness Writes?](https://mlsanity.com/paper/2609.32495/) (2026)
- [DAAF: From Failure Localization to Editable System Assets in LLM Agents](https://mlsanity.com/paper/2609.32498/) (2026)
- [Auditing Agent Actions through Query-Conditioned Attribution](https://mlsanity.com/paper/2609.33676/) (2026)
- [Agent Error Dataset: Scaling 50,000 Error--Diagnosis Pairs for Failure Analysis and Error-Aware Post-Training](https://mlsanity.com/paper/2609.40111/) (2026)
- [Pay for the Fault, Not the Flow: Label-Free In-Flow Multi-Agent Workflow Optimization](https://mlsanity.com/paper/2610.01017/) (2026)
- [It Takes Workflows to Evolve Better Workflows](https://mlsanity.com/paper/2610.01026/) (2026)

## Abstract (arXiv)

Failure attribution, i.e., identifying the responsible agent and decisive step of a failure, is particularly challenging in LLM-based multi-agent systems (MAS) due to their natural-language reasoning, nondeterministic outputs, and intricate interaction dynamics. A reliable benchmark is therefore essential to guide and evaluate attribution techniques. Yet existing benchmarks rely on partially observable traces that capture only agent outputs, omitting the inputs and context that developers actually use when debugging. We argue that failure attribution should be studied under full execution observability, aligning with real-world developer-facing scenarios where complete traces, rather than only outputs, are accessible for diagnosis. To this end, we introduce TraceElephant, a benchmark designed for failure attribution with full execution traces and reproducible environments. We then systematically evaluate failure attribution techniques across various configurations. Specifically, full traces improve attribution accuracy by up to 76% over a partial-observation counterpart, confirming that missing inputs obscure many failure causes. TraceElephant provides a foundation for follow-up failure attribution research, promoting evaluation practices that reflect real-world debugging and supporting the development of more transparent MASs.