Rewards-in-Context: Multi-objective Alignment of Foundation Models with Dynamic Preference Adjustment

• URL: http://arxiv.org/abs/2402.10207v6
• Date: Wed, 16 Oct 2024 03:24:02 GMT
• Title: Rewards-in-Context: Multi-objective Alignment of Foundation Models with Dynamic Preference Adjustment
• Authors: Rui Yang, Xiaoman Pan, Feng Luo, Shuang Qiu, Han Zhong, Dong Yu, Jianshu Chen,
• Abstract summary: We introduce Rewards-in-Context (RiC), which conditions the response of a foundation model on multiple rewards in its prompt context. RiC only requires supervised fine-tuning of a single foundation model and supports dynamic adjustment for user preferences during inference time.
• Score: 46.44464839353993
• License:
• Abstract: We consider the problem of multi-objective alignment of foundation models with human preferences, which is a critical step towards helpful and harmless AI systems. However, it is generally costly and unstable to fine-tune large foundation models using reinforcement learning (RL), and the multi-dimensionality, heterogeneity, and conflicting nature of human preferences further complicate the alignment process. In this paper, we introduce Rewards-in-Context (RiC), which conditions the response of a foundation model on multiple rewards in its prompt context and applies supervised fine-tuning for alignment. The salient features of RiC are simplicity and adaptivity, as it only requires supervised fine-tuning of a single foundation model and supports dynamic adjustment for user preferences during inference time. Inspired by the analytical solution of an abstracted convex optimization problem, our dynamic inference-time adjustment method approaches the Pareto-optimal solution for multiple objectives. Empirical evidence demonstrates the efficacy of our method in aligning both Large Language Models (LLMs) and diffusion models to accommodate diverse rewards with only around 10% GPU hours compared with multi-objective RL baseline.

Related papers

• Adaptive Learning of Design Strategies over Non-Hierarchical Multi-Fidelity Models via Policy Alignment [0.0]
  Multi-fidelity Reinforcement Learning frameworks enhance the efficiency of engineering design by leveraging analysis models with varying levels of accuracy and computational costs. This work proposes ALPHA, a novel multi-fidelity RL framework to efficiently learn a high-fidelity policy by adaptively leveraging an arbitrary set of non-hierarchical, heterogeneous, low-fidelity models alongside a high-fidelity model. The effectiveness of ALPHA is demonstrated in analytical test optimization and octocopter design problems, utilizing two low-fidelity models alongside a high-fidelity one.
  arXiv  Detail & Related papers  (2024-11-16T16:54:33Z)
• TSO: Self-Training with Scaled Preference Optimization [14.3799656174528]
  We propose TSO, a framework for preference optimization that conducts self-training preference learning without training an additional reward model. TSO enhances the diversity of responses by constructing a model matrix and incorporating human preference responses. Experimental results demonstrate that TSO outperforms existing mainstream methods on various alignment evaluation benchmarks.
  arXiv  Detail & Related papers  (2024-08-31T05:37:01Z)
• Joint Demonstration and Preference Learning Improves Policy Alignment with Human Feedback [58.049113055986375]
  We develop a single stage approach named Alignment with Integrated Human Feedback (AIHF) to train reward models and the policy. The proposed approach admits a suite of efficient algorithms, which can easily reduce to, and leverage, popular alignment algorithms. We demonstrate the efficiency of the proposed solutions with extensive experiments involving alignment problems in LLMs and robotic control problems in MuJoCo.
  arXiv  Detail & Related papers  (2024-06-11T01:20:53Z)
• UCB-driven Utility Function Search for Multi-objective Reinforcement Learning [75.11267478778295]
  In Multi-objective Reinforcement Learning (MORL) agents are tasked with optimising decision-making behaviours. We focus on the case of linear utility functions parameterised by weight vectors w. We introduce a method based on Upper Confidence Bound to efficiently search for the most promising weight vectors during different stages of the learning process.
  arXiv  Detail & Related papers  (2024-05-01T09:34:42Z)
• 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)
• Dependency Structure Search Bayesian Optimization for Decision Making Models [29.95525433889418]
  We propose a compact multi-layered architecture modeling the dynamics of agent interactions through the concept of role. We show strong empirical results under malformed or sparse reward.
  arXiv  Detail & Related papers  (2023-08-01T15:56:24Z)
• Exploiting Temporal Structures of Cyclostationary Signals for Data-Driven Single-Channel Source Separation [98.95383921866096]
  We study the problem of single-channel source separation (SCSS) We focus on cyclostationary signals, which are particularly suitable in a variety of application domains. We propose a deep learning approach using a U-Net architecture, which is competitive with the minimum MSE estimator.
  arXiv  Detail & Related papers  (2022-08-22T14:04:56Z)
• Pareto Set Learning for Neural Multi-objective Combinatorial Optimization [6.091096843566857]
  Multiobjective optimization (MOCO) problems can be found in many real-world applications. We develop a learning-based approach to approximate the whole Pareto set for a given MOCO problem without further search procedure. Our proposed method significantly outperforms some other methods on the multiobjective traveling salesman problem, multiconditioned vehicle routing problem and multi knapsack problem in terms of solution quality, speed, and model efficiency.
  arXiv  Detail & Related papers  (2022-03-29T09:26:22Z)
• Efficient Model-Based Multi-Agent Mean-Field Reinforcement Learning [89.31889875864599]
  We propose an efficient model-based reinforcement learning algorithm for learning in multi-agent systems. Our main theoretical contributions are the first general regret bounds for model-based reinforcement learning for MFC. We provide a practical parametrization of the core optimization problem.
  arXiv  Detail & Related papers  (2021-07-08T18:01:02Z)

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.