AI RESEARCH
Beyond Euclidean Proximity: Repairing Latent World Models with Horizon-Matched Trajectory Reachability Metrics
arXiv CS.LG
•
ArXi:2605.22164v1 Announce Type: new Latent world models can contain the state needed for control, yet their terminal-cost interface can expose the planner to the wrong decision-relevant information. In common latent MPC, candidate sequences are ranked by Euclidean distance between predicted terminal and goal latent states; this assumes that raw latent distance weights reachability-relevant variables correctly. We propose trajectory reachability metrics (TRM), a post-hoc terminal-ranking method for fixed latent world models.