paper-with-me

홈 › Papers

Diversity By Design: Leveraging Distribution Matching for Offline Model-Based Optimization

2025-01-30 · Michael S. Yao, James C. Gee, Osbert Bastani

The goal of offline model-based optimization (MBO) is to propose new designs that maximize a reward function given only an offline dataset. However, an important desiderata is to also propose a diverse set of final candidates that capture many optimal and near-optimal design configurations. We propose Diversity in Adversarial Model-based Optimization (DynAMO) as a novel method to introduce design diversity as an explicit objective into any MBO problem. Our key insight is to formulate diversity as a distribution matching problem where the distribution of generated designs captures the inherent diversity contained within the offline dataset. Extensive experiments spanning multiple scientific domains show that DynAMO can be used with common optimization methods to significantly improve the diversity of proposed designs while still discovering high-quality candidates.

📄 PDF Abstract BibTeX arXiv:2501.18768

Code (1)

michael-s-yao/dynamo 공식 구현 pytorch

Tasks

Diversity

Methods 이 논문이 사용한 방법론

SET Dynamic Sparse Training method where weight mask is updated randomly periodically

Similar Papers 제목 키워드 기반

Preference-Guided Diffusion for Multi-Objective Offline Optimization

2025-03-21 · Yashas Annadani, Syrine Belakaria, Stefano Ermon, Stefan Bauer 외

Offline multi-objective optimization aims to identify Pareto-optimal solutions given a dataset of designs and their objective values. In this work, we propose a preference-guided diffusion model that generates Pareto-opt…

Diversity

DISA: Offline Importance Sampling for Distribution-Matching LLM-RL

2026-05-17 · Shaobo Wang, Yujie Chen, Yafeng Sun, Wenjie Qiu 외 arxiv

Modern reasoning agents are increasingly evaluated on their ability to generate multiple valid solution paths, plans, or tool-use traces for a given input. Standard reward-maximizing RL tends to collapse onto the most ea…

SMORE: Score Models for Offline Goal-Conditioned Reinforcement Learning

2023-11-03 · Harshit Sikchi, Rohan Chitnis, Ahmed Touati, Alborz Geramifard 외

Offline Goal-Conditioned Reinforcement Learning (GCRL) is tasked with learning to achieve multiple goals in an environment purely from offline datasets using sparse reward functions. Offline GCRL is pivotal for developin…

Contrastive Learningreinforcement-learningReinforcement LearningRobot Manipulation

Imitation from Observations with Trajectory-Level Generative Embeddings

2026-01-01 · Yongtao Qu, Shangzhe Li, Weitong Zhang arxiv

We consider the offline imitation learning from observations (LfO) where the expert demonstrations are scarce and the available offline suboptimal data are far from the expert behavior. Many existing distribution-matchin…

Diversity-Driven Offline Multi-Objective Optimization via Nested Pareto Set Learning

2026-06-13 · Yiyi Zhu, Yaolin Wen, Xiang Xia, Xin An 외 arxiv

Multi-objective optimization (MOO) has emerged as a powerful approach to solving complex optimization problems involving multiple objectives. In many practical scenarios, function evaluations are unavailable or prohibiti…