Saved in:
Bibliographic Details
Main Authors: Sheebaelhamd, Ziyad, Tschannen, Michael, Muehlebach, Michael, Vernade, Claire
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.14259
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911307408605184
author Sheebaelhamd, Ziyad
Tschannen, Michael
Muehlebach, Michael
Vernade, Claire
author_facet Sheebaelhamd, Ziyad
Tschannen, Michael
Muehlebach, Michael
Vernade, Claire
contents Current transformer-based imitation learning approaches introduce discrete action representations and train an autoregressive transformer decoder on the resulting latent code. However, the initial quantization breaks the continuous structure of the action space thereby limiting the capabilities of the generative model. We propose a quantization-free method instead that leverages Generative Infinite-Vocabulary Transformers (GIVT) as a direct, continuous policy parametrization for autoregressive transformers. This simplifies the imitation learning pipeline while achieving state-of-the-art performance on a variety of popular simulated robotics tasks. We enhance our policy roll-outs by carefully studying sampling algorithms, further improving the results.
format Preprint
id arxiv_https___arxiv_org_abs_2503_14259
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Quantization-Free Autoregressive Action Transformer
Sheebaelhamd, Ziyad
Tschannen, Michael
Muehlebach, Michael
Vernade, Claire
Machine Learning
Robotics
Current transformer-based imitation learning approaches introduce discrete action representations and train an autoregressive transformer decoder on the resulting latent code. However, the initial quantization breaks the continuous structure of the action space thereby limiting the capabilities of the generative model. We propose a quantization-free method instead that leverages Generative Infinite-Vocabulary Transformers (GIVT) as a direct, continuous policy parametrization for autoregressive transformers. This simplifies the imitation learning pipeline while achieving state-of-the-art performance on a variety of popular simulated robotics tasks. We enhance our policy roll-outs by carefully studying sampling algorithms, further improving the results.
title Quantization-Free Autoregressive Action Transformer
topic Machine Learning
Robotics
url https://arxiv.org/abs/2503.14259