The Geometry of Lies: Knowledge and Persona Directions for Detecting Deception in Large Language Models
Md. Abdullah Al Zami, Suraiya Shanaz, Mahdi Hasan, Shoaib Ahmed Dipu, Farig Yousuf Sadeque, Md. Sabbir Ahmed
Abstract
Linear probes are the cheapest LLM lie detectors but often fail to transfer across kinds of lie. Label-free directions from honest and deceptive role-play are untested at 70B scale and on lies about knowledge, where a lie and an honest denial can be identical. We widen a published persona bank with roles that deny knowledge they have, plus honest counterparts. At each model's middle layer the resulting persona direction ranks lies across Liars' Bench better than single-dataset probes on three of its four models; it fails on Qwen, even when Qwen stays in role far more often. On Llama-3.3-70B-Instruct it beats directions from a random persona bank, but its gain over the published direction is within the range of random widenings. On lies about knowledge, persona directions mainly track a reply's wording: where lies and honest denials share their words, our three main contrasts stay inside the random-direction band on Llama and Qwen and reach only its edge on Gemma, though on Mistral one clears it. A knowledge direction contrasts questions the model can answer with ones nobody can, each followed by the same scripted denial. It separates lies from identical honest denials on two models and narrowly on a third, but not on the fourth; its other results partly depend on the benchmark's wording. On the Gender-Secret fine-tune it clears the band by a few hundredths on Llama and fails on Mistral. The sum of the two directions ranks lies better than a supervised probe at the same depth on all four models (on one, both near chance) but is level with the persona direction on Llama's average and behind that probe on held-out lies. More varied role-play questions do not help the persona direction on Liars' Bench.
Create a lesson
Related papers
PlayCraft: Designing Engaging Games from Agent Play Experience
Weihao Cheng
Probing the "Conscious" Mind: Selective Access and Global Workspace in Vision-Language Models
Jiahao Huang, Jun Wang
ProtoDiffusion: Token-per-Layer Generative Modeling of Multi-Layer Images
Adrià Ciurana Lanau, Harleen Hanspal, Lusine Abrahamyan et al.
MM-SystemOneBias: Evaluating Bias in Multimodal Jev-Style System One Decision Models
Jinzhe Li, Qi Wang, Dandan Guo et al.
LLM GerryMandering: Separating Interactions from Additive Effects in LLM Audits
George Rahul, Ronit Jahilal, Bhaskara Hanuma Vedula et al.
GLiDE: A Fast, Calibrated Decision Model Distilled from Reasoning Teachers
Sahibzada Allahyar, Julia White, Dhruv Atreja et al.