RD‑JEPA is a joint‑embedding predictive architecture designed for self‑supervised pretraining on reaction‑diffusion trajectories. The model is pretrained on five parameterized systems and then adapted to three held‑out systems that were not seen during pretraining. Using as few as one, five, or ten complete trajectories from a held‑out system, RD‑JEPA outperforms five supervised surrogate baselines, an independently trained control that removes the trajectory‑dependent predictive latent pathway, and an architecture‑matched model trained from scratch, achieving lower mean relative discrete β field error and mean absolute spatial first‑difference error across various output resolutions, forecast horizons, and adaptation trajectory choices.
arXiv:2603. 02650v2 Announce Type: replace-cross Abstract: Diffusion planners are a strong approach for offline reinforcement learning, but they can fail when value-guided selection favours trajectories that score well yet are locally inconsistent with the environment dynamics, resulting in brittle execution.
arXiv:2609.37113v1 Announce Type: new Abstract: Generalized reaction-diffusion systems encompass diverse transport mechanisms and coupled reaction kinetics. A central question for neural PDE solvers...
ForeDrive introduces a planning-relevant latent world model that is asymmetrically coupled to a Diffusion Transformer planner. The model learns multi‑horizon latent futures with a JEPA‑style world model, while planning gradients update the shared encoder and stop‑gradient routing trains the predictor with forecasting losses only. Gated visual fusion, future‑status injection, and Trajectory‑Adaptive Bias are used to guide trajectory generation without overriding current observations, achieving high performance on NAVSIM benchmarks using only front‑view images and pure imitation learning.
JEPA-x is a cross‑predictive physics grounding method that aligns visual latent dynamics with privileged physical trajectories. By treating visual observations and physical states as two views of the same action‑conditioned trajectory and sharing a predictor, it forces the model to learn a common transition rule for both modalities. The physical branch is only used during training, so deployment incurs no extra cost, and the approach significantly reduces rollout drift and boosts control success across a multi‑task suite.
arXiv:2602. 07339v2 Announce Type: replace Abstract: Diffusion-based trajectory planners can model multi-modal driving behavior, but their iterative denoising process introduces a latency bottleneck for real-time closed-loop deployment.
arXiv:2607. 16090v1 Announce Type: cross Abstract: Transferring policies across domains poses a vital challenge in reinforcement learning, due to the dynamics mismatch between the source and target domains.
arXiv:2501. 14622v5 Announce Type: replace Abstract: Learning efficient representations for decision-making policies is a challenge in imitation learning (IL).
Orthogonal JEPA introduces a latent world‑modeling framework that factorizes predictive states into orthogonal components. By learning basis matrices and dedicated prediction branches, the method reduces redundancy and improves gradient signals for less dominant predictive structures. The factorized states can be synthesized into complete latent representations for downstream tasks such as decoding, planning, or autoregressive rollout, and are evaluated across vision, biology, health, control, and molecular dynamics domains.
arXiv:2602. 04037v3 Announce Type: replace Abstract: Learning domain adaptive policies that can generalize to unseen transition dynamics, remains a fundamental challenge in learning-based control.
Deep Reinforcement Learning (RL) is notoriously sample inefficient. One contributing factor is that RL agents are typically initialized from scratch, forcing them to acquire task-relevant knowledge through online interaction.
The paper introduces an in‑span adaptation technique for reduced‑order models, where the reduced subspace is continually updated using the model’s own predictions via an incremental singular‑value decomposition with a forgetting factor. This creates a trajectory‑informed spectral preconditioner that reweights and realigns the basis without changing the subspace, enabling the model to better absorb future out‑of‑span corrections. The authors demonstrate the method on a 3‑D spiral example and nonlinear PDEs such as viscous Burgers and Fisher–KPP, and relate the approach to in‑context learning in dynamical systems.
arXiv:2606. 20104v1 Announce Type: cross Abstract: Perception for action suggests that representations of the world should be shaped not by visual fidelity alone, but by their relevance for actions.