AD-WM is a new action‑discriminative joint‑embedding world model designed for counterfactual model predictive control. It augments residual latent dynamics with action‑recovery regularization based on inverse dynamics and conditional mutual information, while discarding auxiliary heads at test time so that MPC remains unchanged. Experiments on OGBench‑Cube and other simulation environments show substantial gains in hard‑start success and mean success, and zero‑shot transfer to a Franka robot improves pick‑and‑place success from 42.2% to 71.1%.
By Jiabin Qiu, Zixuan Chen, Hongye Cao, Jieqi Shi, Jing Huo, Yang Gao
The paper introduces a lightweight Fourier auxiliary head to enforce physically-informed structuring of latent states in JEPA-style world models, addressing a newly identified failure mode called physical representation laziness that hampers planning in dynamic environments. Experiments show that this auxiliary supervision improves planning success rates, enhances latent space correlations with key physical properties, and boosts data efficiency, even when the baseline model does not exhibit laziness.
By Penghao Zhu, Salvatore Penachio, Kaustav Mukherjee, Aneesh Jonelagadda
The paper introduces the Dual-Latent World Model (Dual-WM), which separates local execution and long-range planning into distinct latent spaces and dynamics models. A new learning method, Long-Horizon Representation Learning with Weighted Rollout (LoRe), supervises predictions at both levels using exponential horizon weights. Experiments on five goal-conditioned visual control tasks show that Dual-WM improves success rates over strong baselines, especially at longer horizons.
By Delin Zhao, Zhengrong Yue, Shaobin Zhuang, Junlin He, Xiaoyu Chen, Zikang Wang, Yuxin Liu, Limin Wang, Yali Wang
arXiv:2512.03400v2 Announce Type: replace-cross
Abstract: We study how explicit world-modeling objectives affect the internal representations and downstream capability of Transformers, using Rubik's...
By Prakhar Gupta, Henry Conklin, Sarah-Jane Leslie, Andrew Lee
The paper introduces the concept of decision‑metric alignment, which ensures that Euclidean distance to a goal latent in JEPA‑style latent world models correctly ranks action sequences for model‑predictive control. It proposes two metrics—Plan‑Real Spearman and CEM‑stage Spearman—to evaluate latent–real rank agreement, and identifies encoder distortion, terminal rollout error, and candidate margins as key factors affecting alignment. Building on these insights, the authors present DA‑LeWM, an enhanced latent world model that incorporates inverse‑dynamics and demonstration‑conditioned goal‑action heads, leading to faster convergence and higher online success rates compared to the baseline LeWM while maintaining similar probe scores.
By Jiawei Wang, Ke Rui, Yushen Zuo, Yichun Feng, Minglei Li
The paper investigates how latent world models (specifically JEPA-style models) use Euclidean distance to a goal latent as a cost for model‑predictive control (MPC). It introduces two metrics—Plan‑Real Spearman and CEM‑stage Spearman—to evaluate how well latent‑space distances align with real‑task progress, a property termed decision‑metric alignment. By identifying encoder distortion, terminal rollout error, and candidate margins as key factors, the authors propose DA‑LeWM, which augments the base model with inverse‑dynamics and demonstration‑conditioned goal‑action heads, leading to faster convergence and higher online success while maintaining similar probe scores.