Robust Nash Alignment under Preference Uncertainty
cs.AI
Submitted: 2026-09-30
Updated: 2026-09-30
Comments: 38 pages, accepted at 2026 40th Advances in Neural Information Processing System (NeurIPS)
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Computing the optimal distributionally-robust strategy to commit to
- Human Alignment of Large Language Models through Online Preference Optimisation
- MaxMin-RLHF: Alignment with Diverse Human Preferences
- Direct Preference Optimization with Unobserved Preference Heterogeneity: The Necessity of Ternary Preferences
- Stackelberg Self-Annotation: A Robust Approach to Data-Efficient LLM Alignment
- KTO: Model Alignment as Prospect Theoretic Optimization
- Contrastive Preference Learning: Learning from Human Feedback without RL
- Distributionally Robust Reinforcement Learning with Human Feedback
- Optimistic mirror descent in saddle-point problems: Going the extra (gradient) mile
- Achieving Logarithmic Regret in KL-Regularized Zero-Sum Markov Games
- Distributionally Robust Optimization: A Review
- DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter
- Soft Preference Optimization: Aligning Language Models to Expert Distributions
- Generalized Preference Optimization: A Unified Approach to Offline Alignment
- Proximal Point Nash Learning from Human Feedback
- Linear Last-iterate Convergence in Constrained Saddle-point Optimization
- Towards Robust Alignment of Language Models: Distributionally Robustifying Direct Preference Optimization
- Robust LLM Alignment via Distributionally Robust Direct Preference Optimization
- Qwen2 Technical Report
- SLiC-HF: Sequence Likelihood Calibration with Human Feedback
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