EvidenceDemonstrated research proof-of-conceptv1.22.1
In plain English
This page shows what kind of support exists for each claim: real systems, experiments, early evidence, architectural reasoning, open questions, or speculative scenarios.
- Why this matters: AI risk can come from the whole arrangement, not one obvious model.
- What to look for: data, memory, routes, adapters, tools, evaluators, updates, and rollback paths.
- Technical version below: the expert terminology remains available and is linked through the glossary.
On the Resilience of LLM-Based Multi-Agent Collaboration with Faulty Agents
Evidence card
- Claim
- Group behavior and resilience depend on architecture and can differ from isolated-agent behavior.
- Evidence level
- Experimentally observed
- Source
- https://proceedings.mlr.press/v267/huang25ay.html
- Publication date
- 2025-07-13
- Authors or institution
- Jen-Tse Huang, Jiaxu Zhou, Tailin Jin, Xuhui Zhou, Zixi Chen, Wenxuan Wang, Youliang Yuan, Michael R. Lyu, Maarten Sap
- System tested
- LLM-based multi-agent collaboration under faulty-agent conditions.
- Limitations
- Task structures and model choices bound the results.
- What the evidence does show
- Group behavior and resilience depend on architecture and can differ from isolated-agent behavior.
- What the evidence does not show
- That any specific multi-agent production system is safe or unsafe by default.
- Date last reviewed in UTC
- 2026-06-26T00:00:00Z
Site use
This source supports Cognivirus.com pages related to faulty agents, multi-agent collaboration, resilience. Its role is bounded by the limitations listed above.