Scaling Offline Model-Based RL via Jointly-Optimized World-Action Model Pretraining
State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences（多模态人工智能系统国家重点实验室，自动化研究所，中国科学院）；School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院大学人工智能学院）；Alibaba Group（阿里巴巴集团）
Jie Cheng, Ruixi Qiao, Yingwei Ma, Binhua Li, Gang Xiong, Qinghai Miao, Yongbin Li, Yisheng Lv
A significant aspiration of offline reinforcement learning (RL) is to develop a generalist agent with high capabilities from large and heterogeneous datasets. However, prior approaches that scale offline RL either rely heavily on expert trajectories or struggle to generalize to diverse unseen tasks. Inspired by the excellent generalization of world model in conditional video generation, we explore the potential of image observation-based world model for scaling offline RL and enhancing generalization on novel tasks. In this paper, we introduce JOWA: Jointly-Optimized World-Action model, an offline model-based RL agent pretrained on multiple Atari games with 6 billion tokens data to learn general-purpose representation and decision-making ability. Our method jointly optimizes a world-action model through a shared transformer backbone, which stabilize temporal difference learning with large models during pretraining. Moreover, we propose a provably efficient and parallelizable planning algorithm to compensate for the Q-value estimation error and thus search out better policies. Experimental results indicate that our largest agent, with 150 million parameters, achieves 78.9% human-level performance on pretrained games using only 10% subsampled offline data, outperforming existing state-of-the-art large-scale offline RL baselines by 31.6% on averange. Furthermore, JOWA scales favorably with model capacity and can sample-efficiently transfer to novel games using only 5k offline fine-tuning data (approximately 4 trajectories) per game, demonstrating superior generalization. We will release codes and model weights at https://github.com/CJReinforce/JOWA
The recent success of supervised learning methods on ever larger offline datasets has spurred interest in the reinforcement learning (RL) field to investigate whether the same paradigms can be translated to RL algorithms. This research area, known as offline RL, has largely focused on offline policy optimization, aiming to find a return-maximizing policy exclusively from offline data. In this paper, we consider a slightly different approach to incorporating offline data into sequential decision-making. We aim to answer the question, what unsupervised objectives applied to offline datasets are …
King Abdullah University of Science and Technology（阿卜杜拉国王科技大学）；The Swiss AI Lab IDSIA/USI/SUPSI（瑞士人工智能实验室 IDSIA/USI/SUPSI）；NNAISENSE
Offline RL methods have been shown to reduce the need for environment interaction by training agents using offline collected episodes. However, these methods typically require action information to be logged during data collection, which can be difficult or even impossible in some practical cases. In this paper, we investigate the potential of using action-free offline datasets to improve online reinforcement learning, name this problem Reinforcement Learning with Action-Free Offline Pretraining (AFP-RL). We introduce Action-Free Guide (AF-Guide), a method that guides online training by extrac…
A Perspective of Q-value Estimation on Offline-to-Online Reinforcement Learning
The University of Sydney（悉尼大学）；SenseTime Computer Vision Group（商汤科技计算机视觉组）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；CUHK MMLab（香港中文大学多媒体实验室）；Institute for AI, Peking University（北京大学人工智能研究院）
University of California, Berkeley（加州大学伯克利分校）；University of Illinois, Chicago（伊利诺伊大学芝加哥分校）
Our team is proposing to run a full-scale energy demand response experiment in an office building. Although this is an exciting endeavor which will provide value to the community, collecting training data for the reinforcement learning agent is costly and will be limited. In this work, we examine how offline training can be leveraged to minimize data costs (accelerate convergence) and program implementation costs. We present two approaches to doing so: pretraining our model to warm start the experiment with simulated tasks, and using a planning model trained to simulate the real world's reward…
Robo-ValueRL: Reliable Value Estimation for Offline-to-Online Reinforcement Learning
Gaoling School of Artificial Intelligence, Renmin University of China（中国人民大学高瓴人工智能学院）；Beijing Innovation Center of Humanoid Robotics（北京人形机器人创新中心）；Beijing Forestry University（北京林业大学）；Peking University（北京大学）；Microsoft Research（微软研究院）
University of Chinese Academy of Sciences（中国科学院大学）；Peking University（北京大学）；University College London（伦敦大学学院）；Chinese Academy of Sciences（中国科学院）；King’s College London（伦敦国王学院）
The pretrain-finetuning paradigm in large-scale sequence models has made significant progress in natural language processing and computer vision tasks. However, such a paradigm is still hindered by several challenges in Reinforcement Learning (RL), including the lack of self-supervised pretraining algorithms based on offline data and efficient fine-tuning/prompt-tuning over unseen downstream tasks. In this work, we explore how prompts can improve sequence modeling-based offline reinforcement learning (offline-RL) algorithms. Firstly, we propose prompt tuning for offline RL, where a context vec…
Offline reinforcement learning (RL) allows for the training of competent agents from offline datasets without any interaction with the environment. Online finetuning of such offline models can further improve performance. But how should we ideally finetune agents obtained from offline RL training? While offline RL algorithms can in principle be used for finetuning, in practice, their online performance improves slowly. In contrast, we show that it is possible to use standard online off-policy algorithms for faster improvement. However, we find this approach may suffer from policy collapse, whe…
Meta AI Research（Meta人工智能研究院）；University of California, Berkeley（加州大学伯克利分校）；University of California, Los Angeles（加州大学洛杉矶分校）
Recent work has shown that offline reinforcement learning (RL) can be formulated as a sequence modeling problem (Chen et al., 2021; Janner et al., 2021) and solved via approaches similar to large-scale language modeling. However, any practical instantiation of RL also involves an online component, where policies pretrained on passive offline datasets are finetuned via taskspecific interactions with the environment. We propose Online Decision Transformers (ODT), an RL algorithm based on sequence modeling that blends offline pretraining with online finetuning in a unified framework. Our framewor…
GOPlan: Goal-conditioned Offline Reinforcement Learning by Planning with Learned Models
University of Warwick（华威大学）；Hong Kong University of Science and Technology（香港科技大学）；Tsinghua University（清华大学）；University of Cambridge（剑桥大学）；University of Liverpool（利物浦大学）
Université de Montréal（蒙特利尔大学）；Mila（米拉人工智能研究所）；HEC Montréal（蒙特利尔高等商学院）；CIFAR（加拿大高等研究院）；Microsoft Research（微软研究院）
Data efficiency is a key challenge for deep reinforcement learning. We address this problem by using unlabeled data to pretrain an encoder which is then finetuned on a small amount of task-specific data. To encourage learning representations which capture diverse aspects of the underlying MDP, we employ a combination of latent dynamics modelling and unsupervised goal-conditioned RL. When limited to 100k steps of interaction on Atari games (equivalent to two hours of human experience), our approach significantly surpasses prior work combining offline representation pretraining with task-specifi…
Institute of Automation, CAS（中国科学院自动化研究所）；School of Artificial Intelligence, UCAS（中国科学院大学人工智能学院）；Shanghai Jiao Tong University（上海交通大学）；King’s College London（伦敦国王学院）；University College London（伦敦大学学院）
Offline reinforcement learning leverages previously-collected offline datasets to learn optimal policies with no necessity to access the real environment. Such a paradigm is also desirable for multi-agent reinforcement learning (MARL) tasks, given the increased interactions among agents and with the enviroment. Yet, in MARL, the paradigm of offline pre-training with online fine-tuning has not been studied, nor datasets or benchmarks for offline MARL research are available. In this paper, we facilitate the research by providing large-scale datasets, and use them to examine the usage of the Deci…
We present a model-based offline reinforcement learning policy performance lower bound that explicitly captures dynamics model misspecification and distribution mismatch and we propose an empirical algorithm for optimal offline policy selection. Theoretically, we prove a novel safe policy improvement theorem by establishing pessimism approximations to the value function. Our key insight is to jointly consider selecting over dynamics models and policies: as long as a dynamics model can accurately represent the dynamics of the state-action pairs visited by a given policy, it is possible to appro…
Offline reinforcement learning (RL) provides a framework for learning decision-making from offline data and therefore constitutes a promising approach for real-world applications as automated driving. Self-driving vehicles (SDV) learn a policy, which potentially even outperforms the behavior in the sub-optimal data set. Especially in safety-critical applications as automated driving, explainability and transferability are key to success. This motivates the use of model-based offline RL approaches, which leverage planning. However, current state-of-the-art methods often neglect the influence of…
Uni-O4: Unifying Online and Offline Deep Reinforcement Learning with Multi-Step On-Policy Optimization
Shanghai Qi Zhi Institute（上海期智研究院）；Tsinghua University（清华大学）；Shanghai AI Lab（上海人工智能实验室）；The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））
University of Edinburgh（爱丁堡大学）；University College London（伦敦大学学院）；Peking University（北京大学）；BIGAI（北京通用人工智能研究院）
We study a novel setting in offline reinforcement learning (RL) where a number of distributed machines jointly cooperate to solve the problem but only one single round of communication is allowed and there is a budget constraint on the total number of information (in terms of bits) that each machine can send out. For value function prediction in contextual bandits, and both episodic and non-episodic MDPs, we establish information-theoretic lower bounds on the minimax risk for distributed statistical estimators; this reveals the minimum amount of communication required by any offline RL algorit…
Scaling World-Model Reinforcement Learning Through Diffusion Policy Optimization
Dynamic Systems Lab, University College London（伦敦大学学院动态系统实验室）；College of Computing and Data Science, Nanyang Technological University（南洋理工大学计算与数据科学学院）；School of Intelligence Science and Technology, Peking University（北京大学智能科学与技术学院）；Santa Fe Institute（圣塔菲研究所）；School of Statistics and Data Science, Shanghai University of Finance and Economics（上海财经大学统计与数据科学学院）；Department of Computing, Imperial College London（伦敦帝国理工学院计算系）
Reward-free Pretraining for Reinforcement Learning via Occupancy Coverage Maximization
Computational Statistics and Machine Learning - Istituto Italiano di Tecnologia（计算统计与机器学习 - 意大利技术研究院）；AI Centre, Computer Science Department, University College London（人工智能中心，计算机科学系，伦敦大学学院）
Instituto Superior Técnico（高等技术学院）；INESC-ID（葡萄牙国家工程与计算机系统研究所）
We study the interplay between the data distribution and Q-learning-based algorithms with function approximation. We provide a unified theoretical and empirical analysis as to how different properties of the data distribution influence the performance of Q-learning-based algorithms. We connect different lines of research, as well as validate and extend previous results. We start by reviewing theoretical bounds on the performance of approximate dynamic programming algorithms. We then introduce a novel four-state MDP specifically tailored to highlight the impact of the data distribution in the p…
Tencent AI Lab（腾讯AI实验室）；Department of Automation, Tsinghua University（清华大学自动化系）
We study the offline meta-reinforcement learning (OMRL) problem, a paradigm which enables reinforcement learning (RL) algorithms to quickly adapt to unseen tasks without any interactions with the environments, making RL truly practical in many real-world applications. This problem is still not fully understood, for which two major challenges need to be addressed. First, offline RL usually suffers from bootstrapping errors of out-of-distribution state-actions which leads to divergence of value functions. Second, meta-RL requires efficient and robust task inference learned jointly with control p…
