Improving Pre-Trained Vision-Language-Action Policies with Model-Based Search

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Neary, Cyrus, Younis, Omar G., Kuramshin, Artur, Aslan, Ozgur, Berseth, Glen
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918199128227840
author Neary, Cyrus
Younis, Omar G.
Kuramshin, Artur
Aslan, Ozgur
Berseth, Glen
author_facet Neary, Cyrus
Younis, Omar G.
Kuramshin, Artur
Aslan, Ozgur
Berseth, Glen
contents Pre-trained vision-language-action (VLA) models offer a promising foundation for generalist robot policies, but often produce brittle behaviors or unsafe failures when deployed zero-shot in out-of-distribution scenarios. We present Vision-Language-Action Planning & Search (VLAPS) -- a novel framework and accompanying algorithms that embed model-based search into the inference procedure of pre-trained VLA policies to improve their performance on robotic tasks. Specifically, our method biases a modified Monte Carlo Tree Search (MCTS) algorithm -- run using a model of the target environment -- using action priors defined by the VLA policy. By using VLA-derived abstractions and priors in model-based search, VLAPS efficiently explores language-conditioned robotics tasks whose search spaces would otherwise be intractably large. Conversely, by integrating model-based search with the VLA policy's inference procedure, VLAPS yields behaviors that are more performant than those obtained by directly following the VLA policy's action predictions. VLAPS offers a principled framework to: i) control test-time compute in VLA models, ii) leverage a priori knowledge of the robotic environment, and iii) integrate established planning and reinforcement learning techniques into the VLA inference process. Across all experiments, VLAPS significantly outperforms VLA-only baselines on language-specified tasks that would otherwise be intractable for uninformed search algorithms, increasing success rates by as much as 67 percentage points.
format Preprint
id arxiv_https___arxiv_org_abs_2508_12211
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Improving Pre-Trained Vision-Language-Action Policies with Model-Based Search
Neary, Cyrus
Younis, Omar G.
Kuramshin, Artur
Aslan, Ozgur
Berseth, Glen
Robotics
Artificial Intelligence
Pre-trained vision-language-action (VLA) models offer a promising foundation for generalist robot policies, but often produce brittle behaviors or unsafe failures when deployed zero-shot in out-of-distribution scenarios. We present Vision-Language-Action Planning & Search (VLAPS) -- a novel framework and accompanying algorithms that embed model-based search into the inference procedure of pre-trained VLA policies to improve their performance on robotic tasks. Specifically, our method biases a modified Monte Carlo Tree Search (MCTS) algorithm -- run using a model of the target environment -- using action priors defined by the VLA policy. By using VLA-derived abstractions and priors in model-based search, VLAPS efficiently explores language-conditioned robotics tasks whose search spaces would otherwise be intractably large. Conversely, by integrating model-based search with the VLA policy's inference procedure, VLAPS yields behaviors that are more performant than those obtained by directly following the VLA policy's action predictions. VLAPS offers a principled framework to: i) control test-time compute in VLA models, ii) leverage a priori knowledge of the robotic environment, and iii) integrate established planning and reinforcement learning techniques into the VLA inference process. Across all experiments, VLAPS significantly outperforms VLA-only baselines on language-specified tasks that would otherwise be intractable for uninformed search algorithms, increasing success rates by as much as 67 percentage points.
title Improving Pre-Trained Vision-Language-Action Policies with Model-Based Search
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2508.12211