Track manipulation, locomotion and embodied systems.
Search papers, labs, and topics across Lattice.
This paper introduces TierCEM, a novel variant of the Cross-Entropy Method (CEM) that directly incorporates strict constraint priorities into elite selection, bypassing the need for explicit numerical weights. TierCEM sequentially filters sampled candidates based on a priority ordering, and if a constraint eliminates all candidates, it recursively selects elites with minimal violations of that blocking constraint while preserving higher-priority satisfaction. Evaluations on 2D navigation and contact-rich pushing tasks demonstrate that TierCEM effectively respects constraint hierarchies, allowing relaxation of lower-priority constraints when they impede progress towards the primary objective.
Prioritizing task progress over arbitrary weights allows optimization algorithms to intelligently relax conflicting constraints, unlocking solutions previously unattainable.
When constraints conflict, an optimizer must determine which requirements to preserve and which to relax. On the one hand, a priority ordering specifies which requirements take precedence. On the other hand, penalty-based formulations encode their relative importance through numerical weights. Depending on these weights, a solution can improve its weighted score while violating intended priorities. We introduce TierCEM, a variant of the cross-entropy method that incorporates strict constraint priorities directly into elite selection without requiring per-constraint importance weights. TierCEM works by sequentially filtering sampled candidates, from highest- to lowest-priority constraint. If and when a constraint eliminates all remaining candidates, TierCEM returns to the last nonempty set and selects elites with the smallest violations of that blocking constraint, recursively preserving satisfaction of all higher-priority constraints. We evaluate TierCEM on 2D navigation and contact-rich pushing tasks in proprioceptive and learned world-model settings. Experiments show that reversing the constraint ordering changes which constraints are violated under conflict. Prioritizing progress toward the task objective also enables TierCEM to relax lower-priority constraints when they would otherwise prevent further progress.