MAS-PromptBench: When Does Prompt Optimization Improve Multi-Agent LLM Systems?
- Published
- Source
- arXiv
- Paper number
- 475
- Field
- Machine Learning
- arXiv ID
- 2606.23664
Key points
- A benchmark dedicated to prompt optimization for MAS, covering five topologies, three communication protocols, and team sizes from 2 to 10 agents.
- The gains range dramatically from as much as +24 points to as much as -16 points, proving that blind application is risky.
- Optimization works better when communication protocols are more structured: Freeform < Semi-structured < Structured.
- As team size grows, coordination complexity increases and optimization gains shrink, from +2.4 points for 2 agents to -2.1 points on average for 10 agents.
- The natural extension of single-agent optimization tools to MAS is not sufficient, and MAS-specific algorithms are needed.
Paper links
External research summaries. These are not HDATF publications or measured product results.