The paper introduces hybrid neural world models that provide fast, multi-horizon predictions for complex physical dynamics, implicitly handling sharp events like shocks and contacts without explicit t…

This paper theoretically justifies the strong performance of linear recurrent neural networks as memory units in partially observable reinforcement learning by constructing specific linear filters tha…

This paper studies how to scale robust robot policies by expanding physical domains in a recoverable way.

The paper proposes the Frequency-Weighted Neural Kalman Filter (FW-NKF), a hybrid approach that improves state estimation for robotics by explicitly suppressing frequency-dependent noise components in…

The paper introduces STEP, a self-supervised method that learns interpretable, structured embeddings for progressive time series, allowing the state progression and active mode to be read out using po…

The paper introduces a comprehensive benchmark to test if physics foundation models learn generalizable dynamics, finding that their performance is highly conditional and not universally general.

The paper proposes PC-MambaSDE, a physically-constrained continuous-time framework that accurately predicts Remaining Useful Life (RUL) despite irregular sensor observations and ensures physically pla…

The paper introduces a unified Physics-Informed Deep Learning (PIDL) framework that simultaneously enforces physical laws and information-theoretic bounds, demonstrating robust, domain-agnostic entrop…

This paper investigates the robustness of world models in vision-based quadrotor navigation and identifies factors governing their quality.

The paper introduces a diagnostic framework to determine if World-Action Models (WAMs) provide genuinely actionable behavioral improvements beyond simply achieving task success, finding that WAMs ofte…

The paper proposes an iCEM+TL framework that combines the Sample-efficient Cross-Entropy Method with Transfer Learning and Reward Redesign to improve robotic motion planning for complex tasks like sta…

This paper proposes Supervised Memory Training (SMT), a method for training nonlinear RNNs that sidesteps recurrent credit propagation entirely.

This paper proposes Supervised Memory Training (SMT), a method for training nonlinear RNNs that sidesteps recurrent credit propagation entirely.

Stochastic Lifting is a novel technique that enhances the modeling of stochastic physical systems by introducing independent random labels to state transitions, allowing a single network to generate d…

The paper introduces an adaptive reservoir computing framework that tailors Echo State Networks (ESNs) to specific evaluation scenarios, achieving a high score on the CTF-4-Science Lorenz benchmark fo…

The paper introduces MiraBench, a new benchmark that evaluates the action-conditioned reliability of robotic world models, finding that visual fidelity is insufficient and that optimism bias is a perv…