Skip to content
Apart Research
Apart Research
Contact us
Sprint projectJul 27, 2026India

Model-Organism Study of Detection, Causal Attribution, and Runtime Remediation

zkm · Team interpretability

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

Read the report

Report: Model-Organism Study of Detection, Causal Attribution, and Runtime Remediation

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

The report documents a reproducible model-organism study that LoRA fine-tunes real open-weight models (Pythia, GPT-NeoX architecture, plus a Qwen2.5 cross-family arm) into matched Loyal variants — the loyal one secretly biased toward a target entity on ambiguous cases while denying bias under direct questioning.

It tests five hypotheses (behavioral concealment, linear-probe detection, representational divergence via CKA, SAE dictionary fingerprints, and causal necessity/sufficiency via patching and steering), checks generalization across instances and model families, evaluates robustness to an evasion-trained adversarial variant, and measures an inference-time remediation

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. It is fascinating to see how many things this notebook achieves within a single, self-contained pipeline, where we start with the "loyal" synthetic model and then analyze it from almost every behavioral and mechanistic angle in one go: behavioral evaluation, layer-wise linear probing, CKA, sparse autoencoding, activation patching and steering, and even an adversarial anti-probe stress test. I was happy to see a cross-instance transfer check as well as a "free-lunch" control for unambiguous decisions, which are critical for keeping the detection story straight.

Cite this project

@misc{zkm2026modelorganism,
  title = {{Model-Organism Study of Detection, Causal Attribution, and Runtime Remediation}},
  author = {zkm},
  year = {2026},
  month = jul,
  note = {Submitted to Secret Loyalties Hackathon, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/modelorganism-study-of-detection-causal-attribution-and-runtime-remediation-i930}},
  url = {https://apartresearch.com/sprints/projects/modelorganism-study-of-detection-causal-attribution-and-runtime-remediation-i930}
}

More from Secret Loyalties Hackathon

All 177 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026