TD-JEPA: Latent-predictive Representations for Zero-Shot Reinforcement Learning

• URL: http://arxiv.org/abs/2510.00739v1
• Date: Wed, 01 Oct 2025 10:21:18 GMT
• Title: TD-JEPA: Latent-predictive Representations for Zero-Shot Reinforcement Learning
• Authors: Marco Bagatella, Matteo Pirotta, Ahmed Touati, Alessandro Lazaric, Andrea Tirinzoni,
• Abstract summary: We introduce TD-JEPA, which leverages TD-based latent-predictive representations into unsupervised RL.<n> TD-JEPA trains explicit state and task encoders, a policy-conditioned multi-step predictor, and a set of parameterized policies directly in latent space.<n> Empirically, TD-JEPA matches or outperforms state-of-the-art baselines on locomotion, navigation, and manipulation tasks across 13 datasets.
• Score: 63.73629127832652
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Latent prediction--where agents learn by predicting their own latents--has emerged as a powerful paradigm for training general representations in machine learning. In reinforcement learning (RL), this approach has been explored to define auxiliary losses for a variety of settings, including reward-based and unsupervised RL, behavior cloning, and world modeling. While existing methods are typically limited to single-task learning, one-step prediction, or on-policy trajectory data, we show that temporal difference (TD) learning enables learning representations predictive of long-term latent dynamics across multiple policies from offline, reward-free transitions. Building on this, we introduce TD-JEPA, which leverages TD-based latent-predictive representations into unsupervised RL. TD-JEPA trains explicit state and task encoders, a policy-conditioned multi-step predictor, and a set of parameterized policies directly in latent space. This enables zero-shot optimization of any reward function at test time. Theoretically, we show that an idealized variant of TD-JEPA avoids collapse with proper initialization, and learns encoders that capture a low-rank factorization of long-term policy dynamics, while the predictor recovers their successor features in latent space. Empirically, TD-JEPA matches or outperforms state-of-the-art baselines on locomotion, navigation, and manipulation tasks across 13 datasets in ExoRL and OGBench, especially in the challenging setting of zero-shot RL from pixels.

Related papers

• Latent Diffusion Planning for Imitation Learning [78.56207566743154]
  Latent Diffusion Planning (LDP) is a modular approach consisting of a planner and inverse dynamics model.<n>By separating planning from action prediction, LDP can benefit from the denser supervision signals of suboptimal and action-free data.<n>On simulated visual robotic manipulation tasks, LDP outperforms state-of-the-art imitation learning approaches.
  arXiv  Detail & Related papers  (2025-04-23T17:53:34Z)
• Zero-Shot Whole-Body Humanoid Control via Behavioral Foundation Models [71.34520793462069]
  Unsupervised reinforcement learning (RL) aims at pre-training agents that can solve a wide range of downstream tasks in complex environments.<n>We introduce a novel algorithm regularizing unsupervised RL towards imitating trajectories from unlabeled behavior datasets.<n>We demonstrate the effectiveness of this new approach in a challenging humanoid control problem.
  arXiv  Detail & Related papers  (2025-04-15T10:41:11Z)
• Learning from Reward-Free Offline Data: A Case for Planning with Latent Dynamics Models [61.145371212636505]
  Reinforcement learning (RL) learns policies through trial and error, and optimal control, which plans actions using a learned or known dynamics model.<n>We systematically analyze the performance of different RL and control-based methods under datasets of varying quality.<n>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.
  arXiv  Detail & Related papers  (2025-02-20T18:39:41Z)
• Q-value Regularized Transformer for Offline Reinforcement Learning [70.13643741130899]
  We propose a Q-value regularized Transformer (QT) to enhance the state-of-the-art in offline reinforcement learning (RL) QT learns an action-value function and integrates a term maximizing action-values into the training loss of Conditional Sequence Modeling (CSM) Empirical evaluations on D4RL benchmark datasets demonstrate the superiority of QT over traditional DP and CSM methods.
  arXiv  Detail & Related papers  (2024-05-27T12:12:39Z)
• Learning Off-policy with Model-based Intrinsic Motivation For Active Online Exploration [15.463313629574111]
  This paper investigates how to achieve sample-efficient exploration in continuous control tasks. We introduce an RL algorithm that incorporates a predictive model and off-policy learning elements. We derive an intrinsic reward without incurring parameters overhead.
  arXiv  Detail & Related papers  (2024-03-31T11:39:11Z)
• Foundation Policies with Hilbert Representations [54.44869979017766]
  We propose an unsupervised framework to pre-train generalist policies from unlabeled offline data. Our key insight is to learn a structured representation that preserves the temporal structure of the underlying environment. Our experiments show that our unsupervised policies can solve goal-conditioned and general RL tasks in a zero-shot fashion.
  arXiv  Detail & Related papers  (2024-02-23T19:09:10Z)
• Hyperbolic Deep Reinforcement Learning [8.983647543608226]
  We propose a new class of deep reinforcement learning algorithms that model latent representations in hyperbolic space. We empirically validate our framework by applying it to popular on-policy and off-policy RL algorithms on the Procgen and Atari 100K benchmarks.
  arXiv  Detail & Related papers  (2022-10-04T12:03:04Z)

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.