SOURCE-LINKED INTELLIGENCE

The Latent That Never Was: A Forensic Re-run of the CVAE Ablation in Action Chunking Transformers

arXiv · AI, language, vision and robotics · article · Sep 15, 2026 · UTC

Action Chunking Transformers (ACT) are widely used to learn robot manipulation from demonstrations. Their conditional variational autoencoder includes an encoder meant to capture differences between demonstrations during training. The original ACT paper reported that encoder removal dropped the mean success rate from 35% to 2% on two simulated tasks with human demonstrations. We re-ran this ablation in the original code and checked whether the findings depend on the implementation or training data. The published drop does not reappear in our tests, although smaller gains or losses in success r

Read original source ↗ Open in workspace

recordType

paper

region

Global