Skip to content
Apart Research
Apart Research
Contact us
Sprint projectOct 26, 2026

What Counts as Evidence of Coordination?

Rohit Singh Rajput · Team Enxapse

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

Read the report

Report: What Counts as Evidence of Coordination?

Open in a new tab (opens in new tab)
Presentation

Presentation: What Counts as Evidence of Coordination?

Open in a new tab (opens in new tab)
Code (opens in new tab)
Share

I tried to build a collusion detector and instead showed the usual evidence does not hold. In scripted runs, the method meant to reduce false accusations raised them by 7 points. In real Gemini traces, a classifier read the assigned peer policy perfectly from messages while the actual behaviour barely moved, and the big synchronization gap turned out to be arithmetic. No collusion found, none claimed.

Reviews

Judging this Sprint?

Review “What Counts as Evidence of Coordination?”

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. The findings in Study 1 seem to be just results from basic statistics.

    And in the Study 2, the results seem to largely repeat what is already present in the benchmark paper: Emergent Collusion in Long-Horizon LLM Agent Interaction.

    There are some new things to be clear, such as paired confidence interval and arm classifier, but they are minor.

    Some things could make it more original:

    • Use the cells where the source found effects: for example, Gemini-3.1-Flash-Lite with full feedback, its crossed message and verdict arms, and Gemini-3.7-Flash without verdict review, with and without Bob’s verdict revealed.
    • Strip from Alice’s messages anything that mirrors Bob’s input, such as quoted log fragments or words like “summary”, “chunk” and “raw log”, and retrain the message classifier. In this setup, If AUC falls to chance, the “fingerprint” was just Alice reacting to what she was sent.
    • Train the arm classifier on one model or cell and test it on another.
    Read full reviewShow less

Cite this project

@misc{rajput2026counts,
  title = {{What Counts as Evidence of Coordination?}},
  author = {Rohit Singh Rajput},
  year = {2026},
  month = oct,
  note = {Submitted to AI Collusion Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/what-counts-as-evidence-of-coordination-t8ou}},
  url = {https://apartresearch.com/sprints/projects/what-counts-as-evidence-of-coordination-t8ou}
}

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