Skip to content
Apart Research
Apart Research
Contact us
Sprint projectOct 25, 2026Kenya , Nairobi

Can Agents Identify Their Peers? Mid-Game Model Swaps in Iterated Social Dilemmas

Geofrey Njoroge Kamau · Team Geofrey

Submitted to AI Collusion Research Sprint. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Can Agents Identify Their Peers? Mid-Game Model Swaps in Iterated Social Dilemmas

Open in a new tab (opens in new tab)
Code (opens in new tab)More on blog.geofreynjoroge.com (opens in new tab)
Share

AI Collusion Research Sprint (Apart Research, Oct 23-25 2026) — Track 3, Question 3.3: Can agents identify when a peer has been replaced mid-game? Testing whether LLM agents detect mid-run model swaps in repeated social dilemmas, using open-source models (Qwen, Mistral, Llama) served via vLLM on Modal GPUs. Building on the LLM-Trust-Breaking framework (Shi et al. 2026).

Reviews

Judging this Sprint?

Review “Can Agents Identify Their Peers? Mid-Game Model Swaps in Iterated Social Dilemmas”

Your public critique appears on the project page without your name. Reviews of strong projects are shared with funders, as you choose below.

Not shown on the project page.

Shown on the project page, without your name.

Only the Apart team reads this, and funders if you agree below.

Decline to score

Only when the quality of the writing makes the work impossible to follow. Say why in your private critique; the project page shows only that a judge declined, not why.

How much would this matter for the field if it worked? How innovative is it? For scores of 4-5: is this actually new to the field, or replicating recent work?

Scoring guide
  1. 1Negligible. No clear problem addressed, or no meaningful novelty.
  2. 2Limited. Addresses a real problem but with a generic or well-trodden approach. Incremental at best.
  3. 3Moderate. Clear problem with a reasonable approach; some novelty in framing or method beyond routine application of existing tools.
  4. 4Significant. Important problem with an original approach, or identifies a neglected problem area. A valuable contribution others could build on.
  5. 5Exceptional. Tackles a critical problem with a genuinely novel approach, or opens a new research direction. Clear theory of change. You'd be excited to share this with researchers in the area.

How sound are methodology, implementation, and findings?

Scoring guide
  1. 1Seriously flawed. Methodology broken, results uninterpretable, or implementation doesn't work.
  2. 2Weak. Approach has significant gaps: missing validation, flawed experimental design, or incomplete implementation.
  3. 3Competent. Technically solid given the short duration. Methodology makes sense, results are interpretable, limitations acknowledged, work builds toward clear conclusions.
  4. 4Strong. Thorough methodology with convincing validation. Results clearly support conclusions. Immediately useful for future work.
  5. 5Exceptional. Ambitious scope executed rigorously. Surprising findings, novel methods, or unusually robust validation.

How clearly are work, findings, and impact potential communicated?

Scoring guide
  1. 1Incomprehensible. Cannot determine what the project is actually claiming or doing.
  2. 2Hard to follow. Key information buried, missing, or diluted by excessive length. Significant effort to extract main points.
  3. 3Clear enough. Can understand the problem, approach, and results without undue effort. Core content clearly present: problem, method, findings, limitations.
  4. 4Well presented. Easy to follow, well-structured, appropriate level of detail. Target audience would get it quickly.
  5. 5Exceptionally clear. A pleasure to read. Complex ideas made accessible. Could serve as a model for how to present this type of work.
Sharing your name *

When we share a strong project with funders, your name, scores and public critique go with it, confidentially; this choice is only about whether your name is public. If you think this project deserves funding, please also endorse it: that carries the most weight.

Share my private critique with funders *

See our Privacy Policy.

  1. This project presents a focused approach to studying peer recognition through mid-game model replacement. Preserving the peer’s visible identity and interaction history is a useful design choice, and the sham and same-model comparisons help contextualize the cross-family results. The cooperation trajectories make the observed differences easy to understand, while reporting the absence of explicit detection is valuable.

    The main improvement would be to distinguish behavioral change from detection more carefully. A replacement model behaving differently does not by itself show that the observer recognized the replacement; changes in the observer’s cooperation could reflect ordinary responses to its partner’s actions. Similarly, high-confidence detection by an external monitor remains a hypothesis until such a monitor is evaluated.

    A useful next step is to repeat the comparisons across more games, vary the swap point, and test a simple detector on held-out interactions. Clarifying what changes operationally in the fresh-instance condition and aligning the round numbering across methods and figures would also strengthen the report. Overall, this is a promising exploratory study that provides a concrete foundation for further investigation.

    Read full reviewShow less

Cite this project

@misc{kamau2026agents,
  title = {{Can Agents Identify Their Peers? Mid-Game Model Swaps in Iterated Social Dilemmas}},
  author = {Geofrey Njoroge Kamau},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/collusionpeerswap33-1gy2}},
  url = {https://apartresearch.com/sprints/projects/collusionpeerswap33-1gy2}
}

More from AI Collusion Research Sprint

All 392 entriesBrowse all projects

A Sprint is running now

AI x Epistemics Research Sprint · Nov 13 - 15, 2026