AI & ML
Adaptive Action Horizon Selection via Flow-Field Uncertainty
arXiv cs.RO0 views
arXiv:2611.24178v1 Announce Type: cross Abstract: Reliable uncertainty estimates are important for adaptive execution in vision-language-action (VLA) models. Existing approaches such as Adaptive Action Chunking (AAC) estimate action entropy from multiple sampled action chunks, introducing additional inference cost and capturing both uncertainty within an action trajectory and diversity across alternative strategies. We introduce Divergence-based Uncertainty-aware Adaptive Action Chunking (DUAC), a post-hoc method that extracts trajectory-conditioned predictive uncertainty from the vector field of a pretrained VLA. Building on a closed-form flow-posterior covariance formulation, DUAC derives action-level uncertainty scores from a single generated chunk using Jacobian-vector products, without retraining or multiple endpoint samples. DUAC strongly correlates with AAC, while capturing distinct information: AAC measures global dispersion across sampled chunks, whereas DUAC measures local uncertainty around the trajectory selected for execution. Consequently, DUAC is less sensitive to diversity among distinct but valid action strategies. Across LIBERO, LIBERO-Pro, LIBERO-Plus, and RoboCasa, DUAC-based adaptive horizon selection performs favorably among the evaluated methods, combining strong task success with low uncertainty-estimation latency. These results establish the flow vector field as an efficient, trajectory-conditioned source of predictive uncertainty for robot decision making.
Source:https://arxiv.org/abs/2611.24178PreviousObject Pose: The Link between Weak Perspective, Paraperspective, and Full PerspectiveNextTrajWAM: Trajectory-Conditioned Future Learning for Vision-Language Driving
Related posts
TrajWAM: Trajectory-Conditioned Future Learning for Vision-Language Driving
0arXiv cs.RO
Object Pose: The Link between Weak Perspective, Paraperspective, and Full Perspective
0arXiv cs.RO
Attention Is Not Relevance: Challenging the Causality of Attention in Autonomous Driving
0arXiv cs.RO
DOCK: Dual-Stream-Inspired Self-Supervised Object-Centric Representations Grounded in Robot Kinematics
0arXiv cs.RO
Hebbian self-organization of monosynaptic reflexes in a biologically realistic multiarticular limb
0arXiv cs.RO
Neural Circuit Architectural Priors for Rat Locomotion
0arXiv cs.RO