# Testing Interchangeability in LLM Agent Teams

arXiv: https://arxiv.org/abs/2609.05279 · HTML: https://mlsanity.com/paper/2609.05279/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use.
- Builds on 24 of its 37 references that are in the map.
- First posted on arXiv 4 Sep 2026.

## Placed in

- LLM agents and tool use / [General LLM agent frameworks and benchmarks](https://mlsanity.com/agents/general-llm-agent-frameworks-and-benchmarks/)

## Builds on (most cited first, 24 of 24)

- [Generative Agents: Interactive Simulacra of Human Behavior](https://mlsanity.com/paper/2304.03442/) (2023)
- [AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation](https://mlsanity.com/paper/2308.08155/) (2023)
- [MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework](https://mlsanity.com/paper/2308.00352/) (2023)
- [A-MEM: Agentic Memory for LLM Agents](https://mlsanity.com/paper/2502.12110/) (2025)
- [ChatDev: Communicative Agents for Software Development](https://mlsanity.com/paper/2307.07924/) (2023)
- [Why Do Multi-Agent LLM Systems Fail?](https://mlsanity.com/paper/2503.13657/) (2025)
- [On the Utility of Learning about Humans for Human-AI Coordination](https://mlsanity.com/paper/1910.05789/) (2019)
- ["Other-Play" for Zero-Shot Coordination](https://mlsanity.com/paper/2003.02979/) (2020)
- [The Hanabi Challenge: A New Frontier for AI Research](https://mlsanity.com/paper/1902.00506/) (2019)
- [MultiAgentBench: Evaluating the Collaboration and Competition of LLM agents](https://mlsanity.com/paper/2503.01935/) (2025)
- [Collaborating with Humans without Human Data](https://mlsanity.com/paper/2110.08176/) (2021)
- [Emergent social conventions and collective bias in LLM populations](https://mlsanity.com/paper/2410.08948/) (2024)
- [Off-Belief Learning](https://mlsanity.com/paper/2103.04000/) (2021)
- [LLM-Coordination: Evaluating and Analyzing Multi-agent Coordination Abilities in Large Language Models](https://mlsanity.com/paper/2310.03903/) (2023)
- [A Survey of Ad Hoc Teamwork Research](https://mlsanity.com/paper/2202.10450/) (2022)
- [Collab-Overcooked: Benchmarking and Evaluating Large Language Models as Collaborative Agents](https://mlsanity.com/paper/2502.20073/) (2025)
- [A New Formalism, Method and Open Issues for Zero-Shot Coordination](https://mlsanity.com/paper/2106.06613/) (2021)
- [On the Critical Role of Conventions in Adaptive Human-AI Collaboration](https://mlsanity.com/paper/2104.02871/) (2021)
- [Emergent Coordination in Multi-Agent Language Models](https://mlsanity.com/paper/2510.05174/) (2025)
- [Thought Virus: Viral Misalignment via Subliminal Prompting in Multi-Agent Systems](https://mlsanity.com/paper/2603.00131/) (2026)
- [Attractor States Emerge in Multi-Turn LLM Conversations](https://mlsanity.com/paper/2606.30571/) (2026)
- [Sparks of Cooperative Reasoning: LLMs as Strategic Hanabi Agents](https://mlsanity.com/paper/2601.18077/) (2026)
- [Toward an Organizational Science of Multi-Agent LLM Systems: Decoupling Who, How, and Which Algorithm](https://mlsanity.com/paper/2607.25446/) (2026)
- [Bayesian Partner Modelling enables Adaptive Replanning for LLM Coordination](https://mlsanity.com/paper/2608.18490/) (2026)

## Built on by (most cited first, 1 of 1)

- [Copies or Sources? Measuring How LLM Aggregators Count Restated Evidence in Multi-Agent Systems](https://mlsanity.com/paper/2610.06192/) (2026)

## Abstract (arXiv)

Production multi-agent systems replace agents constantly, on the assumption that an agent filling a role is interchangeable with any other agent that can do the job. We test that assumption. Eight teams per setting are formed independently from one base model on the same tasks, each agent keeping a private notebook across ten formation episodes; we then trade role-matched agents between teams and measure what changes on held-out tasks. Against a placebo that reproduces the disruption of a roster change without changing who occupies the seat, a swap costs little in task score but raises the communication a team spends per unit of progress by 16 to 63 percent, and in Hanabi a swapped agent is more expensive than an inexperienced one, consistent with interference from conventions learned with its former partner. In Collab-Overcooked, when the agent that sets the agenda is replaced, most of the extra communication comes from the agent that stayed. Three ablations, over base models, decoding temperature and formation length, move the swap penalty alongside one other quantity: how far independently formed teams drift apart. Greedy decoding lowers both; doubling a team's history raises both. In these settings, agents are more fungible in task outcome than in coordination efficiency, with larger swap effects after longer formation histories.