Article

DEFLECT: Temporal Counterfactual Preference Learning for Delay-Robust Asynchronous VLAs

DEFLECT is an offline post-training framework that turns the temporal gap between prediction and execution into preference supervision, enabling policies to produce actions aligned with the execution-time state even when their inputs are stale.

Part of the event: DEFLECT: Improving Latency Robustness of Asynchronous VLA with Temporal Counterfactual Preference Learning

This page provides a summary and source link only.