[2604.01985] World Action Verifier: Self-Improving World Models via Forward-Inverse Asymmetry

Computer Science > Machine Learning arXiv:2604.01985 (cs) [Submitted on 2 Apr 2026] Title:World Action Verifier: Self-Improving World Models via Forward-Inverse Asymmetry Authors:Yuejiang Liu, Fan Feng, Lingjing Kong, Weifeng Lu, Jinzhou Tang, Kun Zhang, Kevin Murphy, Chelsea Finn, Yilun Du View a PDF of the paper titled World Action Verifier: Self-Improving World Models via Forward-Inverse Asymmetry, by Yuejiang Liu and 8 other authors View PDF HTML (experimental) Abstract:General-purpose world models promise scalable policy evaluation, optimization, and planning, yet achieving the required level of robustness remains challenging. Unlike policy learning, which primarily focuses on optimal actions, a world model must be reliable over a much broader range of suboptimal actions, which are often insufficiently covered by action-labeled interaction data. To address this challenge, we propose World Action Verifier (WAV), a framework that enables world models to identify their own prediction errors and self-improve. The key idea is to decompose action-conditioned state prediction into two factors -- state plausibility and action reachability -- and verify each separately. We show that these verification problems can be substantially easier than predicting future states due to two underlying asymmetries: the broader availability of action-free data and the lower dimensionality of action-relevant features. Leveraging these asymmetries, we augment a world model with (i) a diverse s...