Multi-Objective Exploration and Preference Optimization via Mutual Information

• URL: http://arxiv.org/abs/2607.01392v2
• Date: Fri, 03 Jul 2026 09:34:52 GMT
• Title: Multi-Objective Exploration and Preference Optimization via Mutual Information
• Abstract summary: We propose Multi-Objective Exploration and Optimization Preference via Mutual Information (MI-EPO), an information-theoretic framework.<n>MI-EPO unifies multi-objective exploration and alignment by maximizing the joint conditional mutual information among generated responses, preference feedback, and preference vectors.<n> Experiments on safe alignment and helpful assistant tasks show that MI-EPO significantly improves the alignment between generated responses and preference vectors.
• Score: 28.76886053636383
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Aligning large language models with diverse and heterogeneous human values requires multi-objective alignment methods to effectively trade off conflicting preference dimensions. Current methods achieve this trade-off by training policies conditioned on preference vectors and leveraging online direct preference optimization. However, exploration uncertainty can cause the reward distributions of responses generated under different preference vectors to overlap, and the generated responses may fail to effectively align with the corresponding preference vectors. In this paper, we propose Multi-Objective Exploration and Preference Optimization via Mutual Information (MI-EPO), an information-theoretic framework. It unifies multi-objective exploration and alignment by maximizing the joint conditional mutual information among generated responses, preference feedback, and preference vectors. By incorporating a probabilistic routing mechanism, MI-EPO naturally decomposes objective alignment and preference-aware exploration, encouraging the model to generate responses that are distinguishable and aligned with different preference conditions. Experiments on safe alignment and helpful assistant tasks show that MI-EPO significantly improves the alignment between generated responses and preference vectors, makes the outputs more controllable, and achieves stable trade-offs across multiple objectives.

Related papers

• Meta-Aligner: Bidirectional Preference-Policy Optimization for Multi-Objective LLMs Alignment [35.36029664224222]
  Multi-Objective Alignment aims to align Large Language Models with diverse and often conflicting human values.<n>We propose Meal, i.e., MEta ALigner, a bi-level meta-learning framework enabling bidirectional optimization between preferences and policy responses.
  arXiv  Detail & Related papers  (2026-04-27T08:36:13Z)
• MGDA-Decoupled: Geometry-Aware Multi-Objective Optimisation for DPO-based LLM Alignment [6.301256425456381]
  We introduce MGDA-Decoupled, a geometry-based multi-objective optimisation algorithm.<n>It finds a shared descent direction while explicitly accounting for each objective's convergence dynamics.<n>Experiments on the UltraFeedback dataset show that MGDA-Decoupled achieves the highest win rates against golden responses.
  arXiv  Detail & Related papers  (2026-04-22T15:33:45Z)
• RecNet: Self-Evolving Preference Propagation for Agentic Recommender Systems [109.9061591263748]
  RecNet is a self-evolving preference propagation framework for recommender systems.<n>It proactively propagates real-time preference updates across related users and items.<n>In the backward phase, the feedback-driven propagation optimization mechanism simulates a multi-agent reinforcement learning framework.
  arXiv  Detail & Related papers  (2026-01-29T12:14:31Z)
• Preference-Guided Diffusion for Multi-Objective Offline Optimization [64.08326521234228]
  We propose a preference-guided diffusion model for offline multi-objective optimization.<n>Our guidance is a preference model trained to predict the probability that one design dominates another.<n>Our results highlight the effectiveness of classifier-guided diffusion models in generating diverse and high-quality solutions.
  arXiv  Detail & Related papers  (2025-03-21T16:49:38Z)
• Self-Improvement Towards Pareto Optimality: Mitigating Preference Conflicts in Multi-Objective Alignment [74.25832963097658]
  Multi-Objective Alignment (MOA) aims to align responses with multiple human preference objectives.<n>We find that DPO-based MOA approaches suffer from widespread preference conflicts in the data.
  arXiv  Detail & Related papers  (2025-02-20T08:27:00Z)
• Few-shot Steerable Alignment: Adapting Rewards and LLM Policies with Neural Processes [50.544186914115045]
  Large language models (LLMs) are increasingly embedded in everyday applications.<n> Ensuring their alignment with the diverse preferences of individual users has become a critical challenge.<n>We present a novel framework for few-shot steerable alignment.
  arXiv  Detail & Related papers  (2024-12-18T16:14:59Z)
• Dynamic Detection of Relevant Objectives and Adaptation to Preference Drifts in Interactive Evolutionary Multi-Objective Optimization [2.4374097382908477]
  We study the dynamic nature of DM preferences, which can evolve throughout the decision-making process and affect the relevance of objectives. We propose methods to discard outdated or conflicting preferences when such shifts occur. Our experimental results demonstrate that the proposed methods effectively manage evolving preferences and significantly enhance the quality and desirability of the solutions produced by the algorithm.
  arXiv  Detail & Related papers  (2024-11-07T09:09:06Z)
• Bridging and Modeling Correlations in Pairwise Data for Direct Preference Optimization [75.1240295759264]
  We propose an effective framework for Bridging and Modeling Correlations in pairwise data, named BMC.<n>We increase the consistency and informativeness of the pairwise preference signals through targeted modifications.<n>We identify that DPO alone is insufficient to model these correlations and capture nuanced variations.
  arXiv  Detail & Related papers  (2024-08-14T11:29:47Z)
• Controllable Preference Optimization: Toward Controllable Multi-Objective Alignment [103.12563033438715]
  Alignment in artificial intelligence pursues consistency between model responses and human preferences as well as values. Existing alignment techniques are mostly unidirectional, leading to suboptimal trade-offs and poor flexibility over various objectives. We introduce controllable preference optimization (CPO), which explicitly specifies preference scores for different objectives.
  arXiv  Detail & Related papers  (2024-02-29T12:12:30Z)
• Bayesian Inverse Transfer in Evolutionary Multiobjective Optimization [29.580786235313987]
  We introduce the first Inverse Transfer Multiobjective (invTrEMO) InvTrEMO harnesses the common objective functions in many prevalent areas, even when decision spaces do not precisely align between tasks. InvTrEMO yields high-precision inverse models as a significant byproduct, enabling the generation of tailored solutions on-demand.
  arXiv  Detail & Related papers  (2023-12-22T14:12:18Z)
• Beyond One-Preference-Fits-All Alignment: Multi-Objective Direct Preference Optimization [76.09576643028362]
  We present Multi-Objective Direct Preference Optimization (MODPO) for multiple alignment objectives. MODPO folds language modeling directly into reward modeling, training language models as implicit collective reward models. It theoretically yields the same optimal solutions as MORLHF but is practically more stable and efficient.
  arXiv  Detail & Related papers  (2023-10-05T17:35:26Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.