Atlas / Sociology of AI / Paper

Why Do Multi-Agent LLM Systems Fail?

Swarms & Oversight

Mert Cemri, Melissa Z. Pan, Shuyi Yang, et al. · 2025 · NeurIPS 2025 Datasets & Benchmarks (per listing)

Read the source

Expert coding of traces from seven multi-agent frameworks produced 14 failure modes in three groups: system design, misalignment between agents, and task verification.

Questions it bears on

Help improve this map

Know a paper that belongs here, or think something needs correcting?

Suggest a correction ↗Contributions are reviewed before they appear.