cs.LGNov 12, 2026

A Recipe for Physical Action Model Training

Authors: Shaojiang Wang, Geer Yang

Organizations: Institute of AI for Industries, Chinese Academy of Sciences, Nanjing, China · Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China · University of Chinese Academy of Sciences, Beijing, China

Abstract

Choosing robot model capacity and demonstration budgets also requires decisions about data composition, representation, optimization, and deployment. We connect these decisions through a common conditional-distribution target. Nested learning oracles give exact risk accounting; local information geometry, conditional mutual information, and empirical convergence provide conditional training prescriptions. Known-target numerical checks expose capacity-data coupling, representation-induced mixture mismatch, and an example where deployment KL increases while negative log-likelihood decreases. An analysis of 206 public PushT trajectories evaluates fixed-feature information-design criteria without fitting a policy. Under asymmetric evaluation targets, the selected allocations reduce this criterion by 4.5-6.2% relative to target matching; balanced targets show no benefit. These results clarify which quantities can support training recommendations and when an apparently favorable predictive metric has a different interpretation.

Figures & tables

Explore similar work

CardsList
  1. Robot World Models Are Not Invariant to How the Actions Are Written

    Sep 19, 2026Ahmed Karim, Leon ChlonWorld Models for RoboticsAction-Conditioned World Models

  2. Behavioural Quality Is Not Learning Utility: Objective Mismatch in Human Behaviour Models for Robot Learning

    Oct 26, 2026Haoyang Jiang, Dana Kulic, Michael Burke +1Human Behavior ModelingHuman Behavior Simulation