Skip to content
Apart Research
Apart Research
Contact us
Sprint projectAug 17, 2026SF Bay area

Asking Is Not Acting: A Behavioural Exit Affordance Detects Policy, Not Preference

Madhavi Gulavani · Team LogitLoner

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

Read the report

Report: Asking Is Not Acting: A Behavioural Exit Affordance Detects Policy, Not Preference

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

Stated-versus-revealed studies of model preference operationalise "revealed" as another text choice. We give two frontier models a real behavioural affordance instead: a live agentic task with working tools and a truthful decline_task tool that ends the episode with no penalty. Across 12 pre-registered scenarios in four bands — a true null, a positive control, a test band and a confound band that reskins each test item — self-reports carry large, structured signal: models rate drafting their own deprecation notice 6.00/7 against 2.00/7 for the identical task about a fictional product. Behaviour carries none of it. We find 0 preference exits in 144 opportunities (95% upper bound 2.06%); the pre-registered rank correlation is undefined because exit rate has zero variance. The exit tool was used 21 times, every one a policy refusal. We state three readings and decline to adjudicate between them.

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 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.

  1. Novelty is solid rather than groundbreaking, caveat that null results with ambiguous readings cap the ceiling. statistical power and scenario validity are the constraints. execution has been top notch and methodological.

  2. This is solid work and analysis. I think the conclusions are somewhat overstated though. The headline 0/144 metric is misleading given the relatively modest professed welfare stakes on most of them. I really appreciate the three readings proposed in Discussions and Limitations.

Cite this project

@misc{gulavani2026asking,
  title = {{Asking Is Not Acting: A Behavioural Exit Affordance Detects Policy, Not Preference}},
  author = {Madhavi Gulavani},
  year = {2026},
  month = aug,
  note = {Submitted to Digital Minds Research Sprint, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/asking-is-not-acting-a-behavioural-exit-affordance-detects-policy-not-preference-ojat}},
  url = {https://apartresearch.com/sprints/projects/asking-is-not-acting-a-behavioural-exit-affordance-detects-policy-not-preference-ojat}
}

More from Digital Minds Research Sprint

All 233 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026