# Revealing Hidden Model Behaviors with Task-Specific Self-Reports

arXiv: https://arxiv.org/abs/2607.03640 · HTML: https://mlsanity.com/paper/2607.03640/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 31 references that are in the map.
- First posted on arXiv 3 Jul 2026.

## Builds on (most cited first, 18 of 18)

- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [Representation Engineering: A Top-Down Approach to AI Transparency](https://mlsanity.com/paper/2310.01405/) (2023)
- [Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training](https://mlsanity.com/paper/2401.05566/) (2024)
- [Persona Vectors: Monitoring and Controlling Character Traits in Language Models](https://mlsanity.com/paper/2507.21509/) (2025)
- [Natural Emergent Misalignment from Reward Hacking in Production RL](https://mlsanity.com/paper/2511.18397/) (2025)
- [The Assistant Axis: Situating and Stabilizing the Default Persona of Language Models](https://mlsanity.com/paper/2601.10387/) (2026)
- [Convergent Linear Representations of Emergent Misalignment](https://mlsanity.com/paper/2506.11618/) (2025)
- [Weird Generalization and Inductive Backdoors: New Ways to Corrupt LLMs](https://mlsanity.com/paper/2512.09742/) (2025)
- [NNsight and NDIF: Democratizing Access to Open-Weight Foundation Model Internals](https://mlsanity.com/paper/2407.14561/) (2024)
- [AuditBench: Evaluating Alignment Auditing Techniques on Models with Hidden Behaviors](https://mlsanity.com/paper/2602.22755/) (2026)
- [Eliciting Secret Knowledge from Language Models](https://mlsanity.com/paper/2510.01070/) (2025)
- [Conditional misalignment: common interventions can hide emergent misalignment behind contextual triggers](https://mlsanity.com/paper/2604.25891/) (2026)
- [Introspection Adapters: Training LLMs to Report Their Learned Behaviors](https://mlsanity.com/paper/2604.16812/) (2026)
- [Simple Mechanistic Explanations for Out-Of-Context Reasoning](https://mlsanity.com/paper/2507.08218/) (2025)
- [Censored LLMs as a Natural Testbed for Secret Knowledge Elicitation](https://mlsanity.com/paper/2603.05494/) (2026)
- [Split Personality Training: Revealing Latent Knowledge Through Alternate Personalities](https://mlsanity.com/paper/2602.05532/) (2026)
- [Minimal and Mechanistic Conditions for Behavioral Self-Awareness in LLMs](https://mlsanity.com/paper/2511.04875/) (2025)

## Built on by (most cited first, 1 of 1)

- [A Sense of Proprioception in Language Models](https://mlsanity.com/paper/2610.17495/) (2026)

## Abstract (arXiv)

Fine-tuning can give a language model a hidden behavior--it may give false answers under a narrow condition, or give harmful advice only when a prompt touches a particular topic. We introduce the Stabilized Adapter for self-Report (SAR), a lightweight LoRA adapter that makes a fine-tuned model describe its own hidden behavior in plain language, using only the model and the dataset it was trained on. Across seven implanted behaviors, SAR detects the hidden behavior in every one--even when the model has generalized into broad misalignment that the training data alone does not predict. Introspection Adapters (IA), the closest existing baseline, detects some behaviors from our suite but misses others entirely--and where it misses, it hallucinates, consistently reporting wrong behaviors. SAR retains positive signal on every setting where IA fails and roughly halves the rate of hallucinations. This gives practitioners a more reliable tool to audit a fine-tuned model and answer ``what did it actually learn?'' type of questions.