Geometry Guided Self-Consistency for Physical AI
For physical AI models (e.g., VLAs, WAMs) that generate action trajectories via diffusion, KeyStone provides a practical, training-free inference-time method to improve robustness and success rates.
KeyStone improves task success rates by up to 13.3% over single-trajectory sampling for diffusion-based action generation in physical AI, by drawing K candidate action chunks in parallel, clustering them, and returning the medoid of the largest cluster, with no additional model or wall-clock latency.
State-of-the-art physical AI models generate a chunk of actions per inference through diffusion or flow matching, iteratively refining an initial noise sample into an action trajectory. Because this inference process is inherently stochastic, committing to a single trajectory per round is brittle, and this brittleness compounds across the many sequential rounds that comprise a complete episode. We introduce KeyStone, an inference-time self-consistency method for diffusion-based action generation that draws $K$ candidate action chunks in parallel from a shared model context, clusters them in continuous action space, and returns the medoid of the largest cluster -- no additional model required. Two properties make this practical. First, the compact nature of action trajectories makes diffusion inference memory-bandwidth bound, leaving spare compute capacity to run $K$ chains in parallel with no additional wall-clock latency. Second, unlike token or pixel spaces where distance carries no semantic meaning and selection requires a learned judge, action chunks are geometrically structured such that Euclidean distance directly reflects physical similarity, making selection principled and judge-free. Across diverse vision-language-action models (VLAs) and world-action models (WAMs), KeyStone improves task success rates by up to \textbf{13.3\%} over single-trajectory sampling with negligible latency overhead, while having on par accuracy with model-based selectors at no training cost. We open source KeyStone at https://github.com/dywsjtu/keystone.