Mykel J. Kochenderfer

Uncertainty-aware thresholds slash conjunction screening candidate pairs by up to 71% without missing collisions, outperforming fixed filters and brute-force methods.

Splitting models into fully decoupled, communication-free shards cuts zero-order gradient variance by roughly $1/N$, enabling backprop-free pretraining to beat standard monolithic baselines using significantly fewer perturbations.

Probe-space preconditioning compresses zero-order LLM optimization from 100,000 steps down to 70, outperforming standard backpropagation on OPT-13B while slashing training memory requirements by 90%.

Closed-loop vision verification is no longer bottlenecked by intractable generative surrogates: compact stochastic world models cut parameter counts by 130× while resolving over 80% of previously unverified RGB state spaces.

Achieving a 53% docking success rate with a world model approach, this research outperforms traditional reinforcement learning methods while demonstrating remarkable generalization capabilities.

AST and DiFS reveal critical failure modes in commercial autonomous vehicles that standard simulations overlook, paving the way for enhanced safety measures.