Asymmetries in Spontaneous and Instructed Deception
cs.AI
Submitted: 2026-08-31
Updated: 2026-08-31
Code: https://github.com/JosiahL98/Spont-I
License: http://creativecommons.org/licenses/by/4.0/
The gist: Large language models sometimes deceive users without being instructed to.
Terminology
Abstract
Large language models sometimes deceive users without being instructed to. However, much of the study on deception in models involves instructed deception. We investigated the relationship between instructed and spontaneous (uninstructed) deception in Llama-3.1-70B-Instruct. We compared these two deception settings through direction geometry, cross-setting classifiers, and cross-setting steering. We found the two deception settings share a component of direction (cosine of approximately 0.5) and an asymmetry in the transfer between settings regarding detection and causation. Spontaneous trained classifiers performed better on instructed data than vice versa, and instructed derived directions performed better at steering spontaneous prompts than vice versa. Likewise the best token position to derive steering vectors from differed from the best token position to train and apply classifiers.
Sources
- Understanding intermediate layers using linear classifier probes
- NNsight and NDIF: Democratizing Access to Open-Weight Foundation Model Internals
- Perfect Detection, Failed Control: The Geometry of Knowing vs. Steering in Language Models
- The Llama 3 Herd of Models
- Inference-Time Intervention: Eliciting Truthful Answers from a Language Model
- Pointer Sentinel Mixture Models
- One Probe Won't Catch Them All: Towards Targeted Deception Detection
- AI Deception: A Survey of Examples, Risks, and Potential Solutions
- From Sycophancy to Deception: A Unified Taxonomy for LLM Spontaneous Misalignment
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems