[2506.00751] Alignment Revisited: Are Large Language Models Consistent in Stated and Revealed Preferences?

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs. arXiv Operational Status Get status notifications via email or slack

[2506.00751] Alignment Revisited: Are Large Language Models Consistent in Stated and Revealed Preferences? Skip to main content arXiv is now an independent nonprofit! Learn more × Search arXiv Press Enter to search · Advanced search --> Computer Science > Artificial Intelligence arXiv:2506.00751 (cs) [Submitted on 31 May 2025] Title: Alignment Revisited: Are Large Language Models Consistent in Stated and Revealed Preferences? Authors: Zhuojun Gu , Quan Wang , Shuchu Han View a PDF of the paper titled Alignment Revisited: Are Large Language Models Consistent in Stated and Revealed

related reading

• [2305.18290] Direct Preference Optimization: Your Language Model is Secretly a Reward Modelarxiv.org
• [2503.10990] Statistical Impossibility and Possibility of Aligning LLMs with Human Preferences: From Condorcet Paradox to Nash Equilibriumarxiv.org
• Discovering Language Model Behaviors with Model-Written Evaluations — LessWronglesswrong.com
• [2203.02155] Training language models to follow instructions with human feedbackarxiv.org
• Emotion Concepts and their Function in a Large Language Modeltransformer-circuits.pub
• Externalized reasoning oversight: a research direction for language model alignment — AI Alignment Forumalignmentforum.org
• Interpretable Preferences via Multi-Objective Reward Modeling and Mixture-of-Experts | RLHFlowrlhflow.github.io
• Alignment faking in large language modelsarxiv.org
• [2112.00861] A General Language Assistant as a Laboratory for Alignmentarxiv.org
• Automated Alignment Researchers: Using large language models to scale scalable oversight \ Anthropicanthropic.com
• [2303.17548] Whose Opinions Do Language Models Reflect?arxiv.org
• On the Algorithmic Bias of Aligning Large Language Models with RLHF: Preference Collapse and Matching Regularizationarxiv.org