The paper investigates whether frozen language models can detect a corrupted reward signal by using a single verified record in a two‑option game. In the game, a payout swap and a lying reporter produce identical histories, but a single line confirming the true outcome allows the models to almost perfectly identify the liar. However, the models frequently misclassify honest reporters as liars, with error rates ranging from 26% to 58% depending on model size and wording, indicating a significant limitation in their ability to interpret verified data.
By Arman Nik Khah
arXiv:2609.21996v1 Announce Type: new
Abstract: Large language models can hold knowledge they do not report. A model may sandbag on a capability evaluation, or answer against what it internally knows...
By Hiskias Dingeto
The paper introduces PACT, a method for unlearning deceptive behaviors in large language models by using contrastive forget sets that compare a model’s responses under deceptive and neutral contexts. PACT trains the model to produce pressure‑aware counterfactual targets, preserving benign system‑prompt adherence and reasoning traces while dramatically reducing deception rates from over 50% to under 3% on 32B reasoning models.
By Haoran Tang, Rajiv Khanna
arXiv:2607. 13346v1 Announce Type: cross Abstract: Alignment faking is dangerous because a model can appear compliant under monitoring while preserving behavior it would reveal when unmonitored.
By Aman Mehta
arXiv:2606. 12618v1 Announce Type: new Abstract: Robust lie detectors for language models could enable powerful techniques for auditing, monitoring, and post-hoc investigation of model behaviour, but evaluating them requires testbeds where models verifiably believe the opposite of what they say.
By Alan Cooney, David Africa, Geoffrey Irving
arXiv:2606. 24952v1 Announce Type: cross Abstract: A central aspiration of mechanistic interpretability is controllability: if we know where a behavior is represented in a model's activations, we should be able to modify it.
By Cosimo Galeone, Anna Ettorre, Minsu Park, Giuseppe Ettorre, Daniele Ligorio
The paper introduces the concept of perfect aliasing, where a truth probe that aligns truthful reporting with a task’s prescribed action cannot differentiate between the two based solely on its labels. In a binary reporting game, probes fitted on compliant contexts yield identical optimizations, while on rival contexts their labels are complementary, causing their AUROCs to sum to one across 751 cell-layer pairs. By employing randomized codebooks and mixed-context fitting, the authors demonstrate that separating prescribed output symbols from semantic action enables perfect recovery of truth, achieving an AUROC of 1.000 on rival trials for a reward-trained Gemma-2-9B policy, whereas conventional probes perform near chance.
By Dylan Jayabahu
A 0.6B language model consistently answers YES to 1,200 logical tests, yet its behavior shows no discrimination. Linear probes reveal the correct verdict with high AUC (0.96) and transfer to unseen structures, but a single scalar readout fails due to a saturated decision threshold offset by +4.6 σ. Adjusting this threshold restores behavior accuracy from 50 % to 81 % and improves higher‑scale models, demonstrating that miscalibrated readouts, not hidden knowledge loss, drive performance gaps.
By Gnaneswar Villuri, Hashmath Shaik, Alex Doboli
arXiv:2607.23458v2 Announce Type: replace
Abstract: Chain-of-thought (CoT) explanations support oversight only if they are faithful: the stated reasoning must actually produce the answer. Auditing bl...
By Suramya R. Angdembay, Dikshant Aryal, Nick Rahimi
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.
arXiv:2608. 08881v1 Announce Type: new Abstract: The current work developed seven Retrieval-Augmented Generation (RAG) models based on leading deception theories and compared how deception judgments were made relative to baseline models.
By David M. Markowitz, Timothy R. Levine
arXiv:2607. 20379v1 Announce Type: new Abstract: Natural-language autoencoders score explanations of hidden activations by reconstruction: an explanation is deemed faithful if the activation can be regenerated from it.
By Hiskias Dingeto