Action from Adjacent Set in Physical Space Outperforms the Best Prediction in World Models

Source: arXiv cs.RO·2026-09-01 12:00Publish time unverified

arXiv:2607.23602v2 Announce Type: replace Abstract: Controllers based on sampling and latent world models assign a predicted terminal cost to each candidate action sequence, choose the minimum, execute its first action block, and replan. This rule can fail even when the terminal cost perfectly and accurately reflects the true task objective in the physical world.