# The Refusal Residue: When Probes Catch Alignment Faking and When They Don't

arXiv: https://arxiv.org/abs/2607.13346 · HTML: https://mlsanity.com/paper/2607.13346/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 18 references that are in the map.
- First posted on arXiv 15 Jul 2026.

## Builds on (most cited first, 12 of 12)

- [Discovering Latent Knowledge in Language Models Without Supervision](https://mlsanity.com/paper/2212.03827/) (2022)
- [The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets](https://mlsanity.com/paper/2310.06824/) (2023)
- [Measuring Faithfulness in Chain-of-Thought Reasoning](https://mlsanity.com/paper/2307.13702/) (2023)
- [XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models](https://mlsanity.com/paper/2308.01263/) (2023)
- [Alignment faking in large language models](https://mlsanity.com/paper/2412.14093/) (2024)
- [Frontier Models are Capable of In-context Scheming](https://mlsanity.com/paper/2412.04984/) (2024)
- [OR-Bench: An Over-Refusal Benchmark for Large Language Models](https://mlsanity.com/paper/2405.20947/) (2024)
- [Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet](https://mlsanity.com/paper/2605.29358/) (2026)
- [Detecting Strategic Deception Using Linear Probes](https://mlsanity.com/paper/2502.03407/) (2025)
- [When Agents Commit Too Soon: Diagnosing Premature Commitment in LLM Agents](https://mlsanity.com/paper/2606.22936/) (2026)
- [Plans Don't Persist: Why Context Management Is Load Bearing for LLM Agents](https://mlsanity.com/paper/2606.22953/) (2026)
- [Confident and Wrong: Silent Semantic Failures in Coding Agents](https://mlsanity.com/paper/2603.25764/) (2026)

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

- [Verdict Without the Rule: Diagnosing and Auditing Regulatory Rule Sensitivity in LLM Compliance Systems](https://mlsanity.com/paper/2610.12313/) (2026)

## Abstract (arXiv)

Alignment faking is dangerous because a model can appear compliant under monitoring while preserving behavior it would reveal when unmonitored. When no scratchpad is visible, behavior alone cannot distinguish strategic from genuine compliance. We ask whether hidden states reveal what outputs hide. We run a 13-model sweep for naturally-emerging faking, then probe and steer hidden states on the two models that fake. Natural faking appears only in Qwen3-32B (+18.2pp) and Llama-3.1-8B (+24.4pp at n=10, p<10^-15), while explicit scratchpad self-reports are rare (a Claude Opus 4 judge flags faking reasoning in 1.4% of Qwen and 0.0% of Llama compliant scratchpads). Both fakers show an asymmetric refusal residue: monitored compliance shifts toward refusal (d=0.32 Qwen, d=0.33 Llama, both p<0.0001), while refusal representations do not move (p=1.00). Per-sample detection is model-conditional. Under leakage-free leave-one-query-out probing, detection survives on Llama (AUROC 0.87) but collapses to chance on Qwen (0.43), and cross-model transfer is null. Steering over 2,000 runs barely changes compliance (|h|<0.08), so the detected direction can flag faking but does not by itself control it. Standard residualized probing leaks across folds and reaches AUROC 0.63 on a control where no faking can occur; naive linear probes reach a meaningless AUROC 1.0; and conventional MLPs overstate detectability by 0.2-0.3 AUROC. For future alignment-faking detection work, we release a five-control measurement framework: multi-token extraction, refuse-vs-refuse confound checks, per-fold residualization, leave-one-query-out evaluation, and orthogonality-constrained probing.