Pith. sign in

REVIEW 2 cited by

Jackpot! Alignment as a Maximal Lottery

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2501.19266 v1 pith:6XXGZ2M7 submitted 2025-01-31 cs.AI cs.LGecon.TH

classification cs.AIcs.LGecon.TH
keywords humanmaximalpreferencesrlhfalignmentcitefeedbacklearning
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Reinforcement Learning from Human Feedback (RLHF), the standard for aligning Large Language Models (LLMs) with human values, is known to fail to satisfy properties that are intuitively desirable, such as respecting the preferences of the majority \cite{ge2024axioms}. To overcome these issues, we propose the use of a probabilistic Social Choice rule called \emph{maximal lotteries} as a replacement for RLHF. We show that a family of alignment techniques, namely Nash Learning from Human Feedback (NLHF) \cite{munos2023nash} and variants, approximate maximal lottery outcomes and thus inherit its beneficial properties. We confirm experimentally that our proposed methodology handles situations that arise when working with preferences more robustly than standard RLHF, including supporting the preferences of the majority, providing principled ways of handling non-transitivities in the preference data, and robustness to irrelevant alternatives. This results in systems that better incorporate human values and respect human intentions.

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Shapley-based Data Valuation for LLM Alignment via Sequential Preference Optimization

    cs.LG 2025-12 conditional novelty 5.0 of 10

    Sequential DPO-style training makes coalition policies reconstructable by arithmetic on singleton models, enabling linear-cost approximation of Shapley data values.

  2. Theoretical Tensions in RLHF: Reconciling Empirical Success with Inconsistencies in Social Choice Theory

    stat.ML 2025-06 conditional novelty 5.0 of 10

    RLHF reward modeling satisfies pairwise majority and Condorcet consistency when each response pair is labeled once, because the maximum likelihood ranking then matches the Copeland rule.

Pith tools