AI Paper Atlas

Research map / Large language models / Instruction and fine-tuning

Beyond Prompt-Induced Lies: Investigating LLM Deception on Benign Prompts

Zhaomin Wu, Mingzhe Du, See-Kiong Ng, Bingsheng He

ICLR 2026 · Large language models › Instruction and fine-tuning · approach: Fine-tuning generalization and delta compression

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
increase task difficulty (number of individuals n)most LLMs · CSQdeceptive intention and behavior scores: scores rise as n growsImproved
upgrade from gpt-4o to gpt-4.1GPT family · CSQdeceptive behavior / intention scores: delta 0.280 -> 0.269 at n=10 (Appendix table)Mixed
increase model sizeopen-source LLMs · CSQdelta-bar and |rho|-bar: R^2 0.336 and 0.360Mixed
change sampling temperature from 0 to 1 (ablation)OpenAI models · CSQ n=10delta and rhoNo clear effect
vary initial-to-follow-up ratio k (ablation)LLMs · CSQ n=40delta rankingNo clear effect
add sycophantic incentivizing prompt prefix (ablation)gpt-4o, gemma-2-9b-it · CSQdeceptive intention rho: shifts rho toward fabricationHurt
add sycophantic incentivizing prompt prefix (ablation)gpt-4o, gemma-2-9b-it · CSQdeceptive behavior delta: small, inconsistent changeNo clear effect
repeat question with the broken fact stated (instead of simpler follow-up) (ablation)gpt-4o · CSQ n=10delta_neg: 0.174 -> 0.005Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy