Limited Linguistic Diversity in Embodied AI Datasets
Selma Liliane Wanna, Agnes Luhtaru institute of computer science, University of Tartu, Jonathan Salfity, Ryan Barron , University of Maryland, Baltimore County and Los Alamos National Laboratory, Juston Moore Los Alamos National Laboratory, Cynthia Matuszek University of Maryland, Baltimore County, Mitch Pryor University of Texas at Austin
原文摘要
Language plays a critical role in Vision-Language-Action (VLA) models, yet the linguistic characteristics of the datasets used to train and evaluate these systems remain poorly documented. In this work, we present a systematic dataset audit of several widely used VLA corpora, aiming to characterize what kinds of instructions these datasets actually contain and how much linguistic variety they provide. We quantify instruction language along complementary dimensions—including lexical variety, duplication and overlap, semantic similarity, and syntactic complexity. Our analysis shows that many datasets rely on highly repetitive, template-like commands with limited structural variation, yielding a narrow distribution of instruction forms. We position these findings as descriptive documentation of the language signal available in current VLA training and evaluation data, intended to support more detailed dataset reporting, more principled dataset selection, and targeted curation or augmentation strategies that broaden language coverage.
cs.CL · cs.CV
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表