Learning from Reward-Free Offline Data: A Case for Planning with Latent Dynamics Models

• URL: http://arxiv.org/abs/2502.14819v1
• Date: Thu, 20 Feb 2025 18:39:41 GMT
• Title: Learning from Reward-Free Offline Data: A Case for Planning with Latent Dynamics Models
• Authors: Vlad Sobal, Wancong Zhang, Kynghyun Cho, Randall Balestriero, Tim G. J. Rudner, Yann LeCun,
• Abstract summary: Reinforcement learning (RL) learns policies through trial and error, and optimal control, which plans actions using a learned or known dynamics model. We systematically analyze the performance of different RL and control-based methods under datasets of varying quality. Our results show that model-free RL excels when abundant, high-quality data is available, while model-based planning excels in generalization to novel environment layouts, trajectory stitching, and data-efficiency.
• Score: 31.509112804985133
• License:
• Abstract: A long-standing goal in AI is to build agents that can solve a variety of tasks across different environments, including previously unseen ones. Two dominant approaches tackle this challenge: (i) reinforcement learning (RL), which learns policies through trial and error, and (ii) optimal control, which plans actions using a learned or known dynamics model. However, their relative strengths and weaknesses remain underexplored in the setting where agents must learn from offline trajectories without reward annotations. In this work, we systematically analyze the performance of different RL and control-based methods under datasets of varying quality. On the RL side, we consider goal-conditioned and zero-shot approaches. On the control side, we train a latent dynamics model using the Joint Embedding Predictive Architecture (JEPA) and use it for planning. We study how dataset properties-such as data diversity, trajectory quality, and environment variability-affect the performance of these approaches. Our results show that model-free RL excels when abundant, high-quality data is available, while model-based planning excels in generalization to novel environment layouts, trajectory stitching, and data-efficiency. Notably, planning with a latent dynamics model emerges as a promising approach for zero-shot generalization from suboptimal data.

Related papers

• Off-dynamics Conditional Diffusion Planners [15.321049697197447]
  This work explores the use of more readily available, albeit off-dynamics datasets, to address the challenge of data scarcity in Offline RL. We propose a novel approach using conditional Diffusion Probabilistic Models (DPMs) to learn the joint distribution of the large-scale off-dynamics dataset and the limited target dataset.
  arXiv  Detail & Related papers  (2024-10-16T04:56:43Z)
• SeMOPO: Learning High-quality Model and Policy from Low-quality Offline Visual Datasets [32.496818080222646]
  We propose a new approach to model-based offline reinforcement learning. We provide a theoretical guarantee of model uncertainty and performance bound of SeMOPO. Experimental results show that our method substantially outperforms all baseline methods.
  arXiv  Detail & Related papers  (2024-06-13T15:16:38Z)
• Distributional Successor Features Enable Zero-Shot Policy Optimization [36.53356539916603]
  This work proposes a novel class of models, i.e., Distributional Successor Features for Zero-Shot Policy Optimization (DiSPOs) DiSPOs learn a distribution of successor features of a stationary dataset's behavior policy, along with a policy that acts to realize different successor features achievable within the dataset. By directly modeling long-term outcomes in the dataset, DiSPOs avoid compounding error while enabling a simple scheme for zero-shot policy optimization across reward functions.
  arXiv  Detail & Related papers  (2024-03-10T22:27:21Z)
• Deep autoregressive density nets vs neural ensembles for model-based offline reinforcement learning [2.9158689853305693]
  We consider a model-based reinforcement learning algorithm that infers the system dynamics from the available data and performs policy optimization on imaginary model rollouts. This approach is vulnerable to exploiting model errors which can lead to catastrophic failures on the real system. We show that better performance can be obtained with a single well-calibrated autoregressive model on the D4RL benchmark.
  arXiv  Detail & Related papers  (2024-02-05T10:18:15Z)
• MOTO: Offline Pre-training to Online Fine-tuning for Model-based Robot Learning [52.101643259906915]
  We study the problem of offline pre-training and online fine-tuning for reinforcement learning from high-dimensional observations. Existing model-based offline RL methods are not suitable for offline-to-online fine-tuning in high-dimensional domains. We propose an on-policy model-based method that can efficiently reuse prior data through model-based value expansion and policy regularization.
  arXiv  Detail & Related papers  (2024-01-06T21:04:31Z)
• Data-Efficient Task Generalization via Probabilistic Model-based Meta Reinforcement Learning [58.575939354953526]
  PACOH-RL is a novel model-based Meta-Reinforcement Learning (Meta-RL) algorithm designed to efficiently adapt control policies to changing dynamics. Existing Meta-RL methods require abundant meta-learning data, limiting their applicability in settings such as robotics. Our experiment results demonstrate that PACOH-RL outperforms model-based RL and model-based Meta-RL baselines in adapting to new dynamic conditions.
  arXiv  Detail & Related papers  (2023-11-13T18:51:57Z)
• Learning Objective-Specific Active Learning Strategies with Attentive Neural Processes [72.75421975804132]
  Learning Active Learning (LAL) suggests to learn the active learning strategy itself, allowing it to adapt to the given setting. We propose a novel LAL method for classification that exploits symmetry and independence properties of the active learning problem. Our approach is based on learning from a myopic oracle, which gives our model the ability to adapt to non-standard objectives.
  arXiv  Detail & Related papers  (2023-09-11T14:16:37Z)
• Reparameterized Policy Learning for Multimodal Trajectory Optimization [61.13228961771765]
  We investigate the challenge of parametrizing policies for reinforcement learning in high-dimensional continuous action spaces. We propose a principled framework that models the continuous RL policy as a generative model of optimal trajectories. We present a practical model-based RL method, which leverages the multimodal policy parameterization and learned world model.
  arXiv  Detail & Related papers  (2023-07-20T09:05:46Z)
• Model-Based Reinforcement Learning with Multi-Task Offline Pretraining [59.82457030180094]
  We present a model-based RL method that learns to transfer potentially useful dynamics and action demonstrations from offline data to a novel task. The main idea is to use the world models not only as simulators for behavior learning but also as tools to measure the task relevance. We demonstrate the advantages of our approach compared with the state-of-the-art methods in Meta-World and DeepMind Control Suite.
  arXiv  Detail & Related papers  (2023-06-06T02:24:41Z)
• Evaluating model-based planning and planner amortization for continuous control [79.49319308600228]
  We take a hybrid approach, combining model predictive control (MPC) with a learned model and model-free policy learning. We find that well-tuned model-free agents are strong baselines even for high DoF control problems. We show that it is possible to distil a model-based planner into a policy that amortizes the planning without any loss of performance.
  arXiv  Detail & Related papers  (2021-10-07T12:00:40Z)

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.