Topological Feasibility Guarantees for Differentiable Predictive Control
arXiv:2608.10332
2026
Dynamics
2 ideas extracted · analyzed Sep 1, 2026
What the math gives to ML
The paper offers a transferable mechanism for converting model-based differentiable policies into deterministically feasible controllers: embed differentiable dynamics in the policy computation graph, train with Control Barrier Function inequalities, and exploit topological and geometric properties of the reachable safe set. The key asset is a finite-sample-to-continuous-domain argument: if sampled closed-loop trajectories cover the relevant reachable set and satisfy a strict safety margin, Lipschitz continuity can certify safety between samples. For neural networks, the most promising implementation is a CBF-constrained policy or neural world-model controller with adaptive trajectory sampling, accompanied by estimates of the safety margin, residual Lipschitz constant, and coverage radius.
Ideas from this paper
✗ Failed on benchmark
2026
Train a neural policy through a differentiable dynamics model while enforcing a Control Barrier Function condition at every rollout state, rather than applying a penalty only to observed constraint violations. The barrier residual becomes a local certificate that the learned policy points inward at the boundary of the safe set, allowing safety to be checked on unseen states when combined with a margin and Lipschitz bound.
Useful8/10
Difficulty5/10
Novelty5/10
△ Mechanism confirmed, baseline not beaten
2026
Use the reachable-safe-set viewpoint to make training data generation adaptive: maintain an approximation of the states reached by the current neural policy, identify boundary regions with weak barrier margin, and sample there until the set is sufficiently covered. This replaces random rollout expansion with a measurable coverage condition that can support finite-sample safety claims.
Useful7/10
Difficulty6/10
Novelty7/10