Theme of inquiry

What causes coordination failures and safety problems in multi-agent systems?

A question within its area, explored through 5 lines of inquiry below — each a family of specific questions the research asks.


When do multi-agent systems outperform single frontier models?

66 specific questions

See all 66 questions in this line of inquiry
When do multi-agent systems provide sufficient quality returns on token investment?

19 specific questions

See all 19 questions in this line of inquiry
Can multi-agent systems avoid converging on false agreement without deliberation?

66 specific questions

See all 66 questions in this line of inquiry
How do multi-agent LLM systems fail distinctly compared to single agents?

41 specific questions

See all 41 questions in this line of inquiry
How do standardized protocols improve multi-agent coordination and reliability?

63 specific questions

See all 63 questions in this line of inquiry