BWhat changed
Generated the next diffusion/flow action chunk while executing the current one, freezing already committed actions and inpainting the remaining sequence to reduce latency-induced discontinuities.
Scientific scope
Generated the next diffusion/flow action chunk while executing the current one, freezing already committed actions and inpainting the remaining sequence to reduce latency-induced discontinuities. Action chunking, flow/diffusion and inpainting are inherited. No-retraining applies to compatible policy/runtime assumptions, not all robot controllers or arbitrary inference delays. Later training-time conditioning is a distinct modification; empirical smoothness is not a safety/stability proof.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Action chunking, flow/diffusion and inpainting are inherited. No-retraining applies to compatible policy/runtime assumptions, not all robot controllers or arbitrary inference delays. Later training-time conditioning is a distinct modification; empirical smoothness is not a safety/stability proof.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
- Real-Time Execution of Action Chunking Flow Policies
- Black et al.: asynchronous action chunks with committed-prefix inpainting — institutional account
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction