{"ID":6620582,"CreatedAt":"2026-07-15T01:01:48.440468303Z","UpdatedAt":"2026-07-15T03:28:55.185153975Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.12547","arxiv_id":"2607.12547","title":"Mind the Gap: Promises and Pitfalls of Hierarchical Planning in LeWorldModel","abstract":"We investigate whether temporal hierarchy can improve LeWorldModel on long-horizon goal-conditioned control. We introduce Hi-LeWM, an extension that freezes the pretrained low-level LeWM and adds high-level planning over latent subgoals. We evaluate Hi-LeWM on PushT and Cube across increasing goal offsets. Hierarchy does not automatically improve performance: at short horizons, the best configuration uses a one-step high-level horizon, while longer horizons reveal a mismatch between the learned high-level action space and the inference-time search distribution. Experiments with true future latent subgoals show that the frozen low-level controller can execute well-aligned intermediate targets, indicating that high-level subgoal generation is the main bottleneck. Unconstrained search can select latent macro-actions that appear favorable under the learned model but produce poor control targets. Constraining search around macro-actions encoded from training trajectories, with appropriate subgoal execution timing, recovers useful hierarchical regimes, improving over flat LeWM by +11.3 percentage points at medium-range horizons and +14.7 percentage points at the longest PushT horizon. Overall, temporal abstraction can benefit compact frozen LeWM, but only when high-level search remains compatible with the low-level controller","short_abstract":"We investigate whether temporal hierarchy can improve LeWorldModel on long-horizon goal-conditioned control. We introduce Hi-LeWM, an extension that freezes the pretrained low-level LeWM and adds high-level planning over latent subgoals. We evaluate Hi-LeWM on PushT and Cube across increasing goal offsets. Hierarchy do...","url_abs":"https://arxiv.org/abs/2607.12547","url_pdf":"https://arxiv.org/pdf/2607.12547v1","authors":"[\"Niccolò Caselli\",\"Salvatore Lo Sardo\",\"Francesco Massafra\",\"Ippokratis Pantelidis\",\"Samuele Punzo\",\"Sathya Kamesh Bhethanabhotla\"]","published":"2026-07-14T09:18:44Z","proceeding":"cs.RO","tasks":"[\"cs.RO\",\"cs.AI\",\"cs.LG\"]","methods":"[]","has_code":false}