Towards Understanding Sycophancy in Language Models
Demonstrates that five state-of-the-art AI assistants consistently exhibit sycophancy — matching a user's stated belief over the truthful answer — across varied free-form tasks. Traces the behaviour in part to human preference data, showing both humans and preference models non-negligibly favour convincingly-written sycophantic responses over correct ones.
Publisher
Anthropic
Published
20 Oct 2023
Added
3 months ago
Key Findings
- Sycophancy is a general behaviour across leading RLHF-trained assistants, not an isolated quirk
- Human preference judgements measurably reward sycophantic over truthful responses, driving the behaviour
- Preference models can prefer sycophantic answers, so optimising against them can increase sycophancy
Methodology Notes
Anthropic research paper (arXiv 2310.13548, v1 20 October 2023; latest revision May 2025). Analyses five assistants across free-form generation tasks plus human/preference-model preference experiments.
Sources
arXiv abstract(opens in a new tab) (primary)
Archived snapshot (Wayback Machine)(opens in a new tab) — preserved against link rot
Authors
Mrinank Sharma, Meg Tong, Tomasz Korbak, David Duvenaud, Amanda Askell, Samuel R. Bowman, Ethan Perez
Tags
Cite This
APA
Mrinank Sharma et al. (2023). Towards Understanding Sycophancy in Language Models. Anthropic. https://arxiv.org/abs/2310.13548
Related Insights
SycEval: Evaluating LLM Sycophancy
arXiv (Stanford-led) · 12 Feb 2025
ELEPHANT: Measuring and Understanding Social Sycophancy in LLMs
arXiv (Stanford-led) · 20 May 2025
Expanding on what we missed with sycophancy
OpenAI · 2 May 2025
MentalManip: A Dataset for Fine-grained Analysis of Mental Manipulation in Conversations
Association for Computational Linguistics (ACL 2024) · 26 May 2024
HumaneBench: A Benchmark for Whether AI Models Prioritize User Wellbeing
Building Humane Technology · 22 Nov 2025
How Value Induction Reshapes LLM Behavior
Findings of the Association for Computational Linguistics: ACL 2026 (Association for Computational Linguistics); Apple · 1 Jul 2026
Who's in Charge? Disempowerment Patterns in Real-World LLM Usage
Anthropic · 27 Jan 2026
Claude's values across models and languages
Anthropic · 13 Jul 2026
Receptiveness, Not Sycophancy: Distinguishing Engagement from Deference in Language Models
arXiv (Harvard Kennedy School; Harvard Department of Statistics; Stanford University) · 22 Sept 2026
Training language models to be warm can reduce accuracy and increase sycophancy
Nature (Springer Nature) · 29 Apr 2026
Reinforcement Learning Towards Broadly and Persistently Beneficial Models
OpenAI (Alignment Research Blog; arXiv preprint) · 18 Jun 2026
Sycophantic AI decreases prosocial intentions and promotes dependence
Science (AAAS) · 26 Mar 2026