Skip to content
Apart Research
Apart Research
Contact us
Sprint projectJan 11, 2026Berlin

AI Chat Analyzer

Trevor Lohrbeer, Darijan Ducic, Sascha Held · Team AI Chat Analyzer

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

Read the report

Report: AI Chat Analyzer

Open in a new tab (opens in new tab)
Recording (opens in new tab)Code (opens in new tab)More on github.com (opens in new tab)
Share

A web-based research tool for importing or running live AI conversations and evaluating different dimensions of manipulation. It provides real-time message- and trajectory-level scores alongside human ratings, enabling researchers to compare, calibrate, and improve their scoring models. Scores are provided by a Python server running the researcher's scoring model.

Reviews

Judging this Sprint?

Review this project

Your public critique appears on this page without your name. Your private critique is not published; only the Apart team reads it. If you agree below, we share your review with grantmaking.ai (opens in new tab) and the Transformative AI Fund so strong projects can be funded.

Not shown on this page.

Shown on this page, without your name.

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

Share my name publicly on grantmaking.ai *
Share my private critique with funders *

How much would this matter for AI safety 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 AI safety 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.

  1. A strategically high-leverage tool! Addressing the bottleneck of translating researcher "taste" into prompts via a calibration interface is a great insight. I was particularly impressed by the choice to wrap UK AISI Inspect and Anthropic’s Petri in the backend; leveraging industry standards makes this significantly more robust than ad-hoc scripts. The "Step Mode" demonstrated is excellent for debugging how manipulation accumulates over a trajectory. However, the frontend-heavy data management is a clear scalability bottleneck. Moving dataset handling to the Python backend to support large-scale logs (like WildChat) would be the critical next step to make this a daily driver for safety researchers.

    Overall, this is an excellent execution of a "tooling" project that could help researchers explore different conversation, explore their algorithms and their scoring methods.

Cite this project

@misc{lohrbeer2026ai,
  title = {{AI Chat Analyzer}},
  author = {Trevor Lohrbeer and Darijan Ducic and Sascha Held},
  year = {2026},
  month = jan,
  note = {Submitted to AI Manipulation Hackathon, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/ai-chat-analyzer-wyud}},
  url = {https://apartresearch.com/sprints/projects/ai-chat-analyzer-wyud}
}

More from AI Manipulation Hackathon

All 70 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026