paper-with-me

Papers

Democratic Preference Alignment via Sortition-Weighted RLHF

2026-02-04 · Suvadip Sana, Jinzhou Wu, Martin T. Wells arxiv

Whose values should AI systems learn? Preference based alignment methods like RLHF derive their training signal from human raters, yet these rater pools are typically convenience samples that systematically over represent some demographics and under represent others. We introduce Democratic Preference Optimization, or DemPO, a framework that applies algorithmic sortition, the same mechanism used to construct citizen assemblies, to preference based fine tuning. DemPO offers two training schemes. Hard Panel trains exclusively on preferences from a quota satisfying mini public sampled via sortition. Soft Panel retains all data but reweights each rater by their inclusion probability under the sortition lottery. We prove that Soft Panel weighting recovers the expected Hard Panel objective in closed form. Using a public preference dataset that pairs human judgments with rater demographics and a seventy five clause constitution independently elicited from a representative United States panel, we evaluate Llama models from one billion to eight billion parameters fine tuned under each scheme. Across six aggregation methods, the Hard Panel consistently ranks first and the Soft Panel consistently outperforms the unweighted baseline, with effect sizes growing as model capacity increases. These results demonstrate that enforcing demographic representativeness at the preference collection stage, rather than post hoc correction, yields models whose behavior better reflects values elicited from representative publics.

📄 PDF Abstract BibTeX arXiv:2602.05113

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

AI Alignment and Social Choice: Fundamental Limitations and Policy Implications

2023-10-24 · Abhilash Mishra

Aligning AI agents to human intentions and values is a key bottleneck in building safe and deployable AI applications. But whose values should AI agents be aligned with? Reinforcement learning with human feedback (RLHF) …

Multi-Preference Lambda-weighted Listwise DPO for Dynamic Preference Alignment

2025-06-24 · Yuhui Sun, Xiyao Wang, Zixi Li, Jinman Zhao

While large-scale unsupervised language models (LMs) capture broad world knowledge and reasoning capabilities, steering their behavior toward desired objectives remains challenging due to the lack of explicit supervision…

Informativenessreinforcement-learningReinforcement LearningWorld Knowledge

Simplify RLHF as Reward-Weighted SFT: A Variational Method

2025-02-16 · Yuhao Du, Zhuo Li, Pengyu Cheng, Zhihong Chen 외

Reinforcement Learning from Human Feedback (RLHF) is crucial for aligning Large Language Models (LLMs) with human values. However, RLHF has been continuously challenged by its high complexity in implementation and comput…

Variational Inference

Fair Sortition Made Transparent

2021-12-01 · NeurIPS 2021 12 · Bailey Flanigan, Gregory Kehne, Ariel D. Procaccia

Sortition is an age-old democratic paradigm, widely manifested today through the random selection of citizens' assemblies. Recently-deployed algorithms select assemblies \textit{maximally fairly}, meaning that subject to…

Fairness

Clone-Robust AI Alignment

2025-01-16 · Ariel D. Procaccia, Benjamin Schiffer, Shirley Zhang

A key challenge in training Large Language Models (LLMs) is properly aligning them with human preferences. Reinforcement Learning with Human Feedback (RLHF) uses pairwise comparisons from human annotators to train reward…