LA4VLA: Learning to Act without Seeing via Language-Action Pretraining

• URL: http://arxiv.org/abs/2606.27295v2
• Date: Fri, 26 Jun 2026 13:05:58 GMT
• Title: LA4VLA: Learning to Act without Seeing via Language-Action Pretraining
• Abstract summary: Vision-Language-Action (VLA) models are commonly pretrained on robot demonstrations by jointly mapping visual observations and language instructions to actions.<n>We propose LA4VLA, a language-action pretraining framework that enables policies to acquire language-conditioned action priors without visual observations.
• Score: 51.209359878543005
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Vision-Language-Action (VLA) models are commonly pretrained on robot demonstrations by jointly mapping visual observations and language instructions to actions. However, dense visual-action supervision can dominate the comparatively sparse language-action signal. As a result, policies may rely on visual shortcuts rather than learn how language conditions action execution, making them sensitive to visual variations. To address this limitation, we propose LA4VLA, a language-action pretraining framework that enables policies to acquire language-conditioned action priors without visual observations. These priors capture reusable manipulation skills shared across tasks and scenes, reducing reliance on scene-specific visual cues. Specifically, LA4VLA decomposes expert demonstration trajectories into atomic action segments and pairs each segment with a corresponding low-level action description. This yields LA-33K, a dataset of 33K Language-Action (LA) episodes derived entirely from existing demonstrations without additional robot data collection. We further develop LA4VLA-1B, a lightweight 1B-parameter VLA model, and investigate three paradigms for incorporating language-action supervision into VLA learning: LA-only pretraining, sequential LA-to-VLA pretraining, and mixed LA-VLA pretraining. Across simulation and real-world tasks, LA-pretrained policies consistently outperform matched VLA-pretrained counterparts, while combining LA and VLA supervision leads to further gains. In particular, mixed LA-VLA pretraining improves the average success rate of LA4VLA-1B over the no-pretraining baseline by up to 17.8 and 45.0 percentage points in simulation and real-world tasks, respectively. These results establish LA4VLA as an effective and complementary pretraining strategy for building stronger and more robust VLA policies.

Related papers

• CLAP: Direct VLM-to-VLA Adaptation via Language-Action Grounding [16.03159066719039]
  CLAP (Causal Language-Action Prediction) prepends each numeric action sequence with a natural-language action description.<n>We will release CLAP at 0.8B, 2B, and 4B as an open-weight, multi-scale compact VLA family from a single VLM lineage.
  arXiv  Detail & Related papers  (2026-07-09T22:34:11Z)
• APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies [22.87409999086972]
  We propose a two-stage training method emphasizing Action expert PreTraining.<n>In Stage 1, the action expert is pretrained as a VA prior on vision-action pairs from a frozen VLM, bypassing the language imbalance.<n>In Stage 2, language tokens are injected through a gated fusion mechanism that integrates VLM features while preserving the learned visuomotor prior.
  arXiv  Detail & Related papers  (2026-06-10T17:34:25Z)
• Two Bridges, One Pathway: From VLMs to Generalizable VLAs with Embodied Trajectory-Coupled Data [93.7685703383343]
  Vision-language models (VLMs) are powerful general-purpose reasoners, yet converting them into robot control policies is surprisingly difficult.<n>We argue that this gap can be bridged gradually with the right intermediate data.<n>We show that mixing task-relevant out-of-distribution ETC data with a small amount of action data enables the model to generalize to novel visual-language conditions.
  arXiv  Detail & Related papers  (2026-06-07T08:57:51Z)
• LARA: Latent Action Representation Alignment for Vision-Language-Action Models [21.4758859185298]
  Latent Action Models (LAM) learn latent action representations from visual dynamics to provide additional supervision for VLA learning.<n>We propose Latent Action Representation Alignment (LARA), a plug-and-play framework that jointly optimize LAM and VLA via representation alignment.<n>We demonstrate LARA versatility and effectiveness for pre-training, post-training enhancement of pre-trained VLA models, and LAM refinement.
  arXiv  Detail & Related papers  (2026-06-05T09:51:25Z)
• QuoVLA: Quotient Space for Vision-Language-Action Models [51.02329790939691]
  Vision-Language-Action (VLA) models commonly adapt pretrained VLM latents to robot control by mapping visual observations and language instructions to continuous actions.<n>Our textitQuotient Theory for VLA shows that pretrained VLM latents are not action-insufficient but action-sufficient.<n>We propose QuoVLA, a quotient-space framework for VLA that compresses pretrained VLM latents into action-sufficient representations.
  arXiv  Detail & Related papers  (2026-05-24T06:28:53Z)
• InstructVLA: Vision-Language-Action Instruction Tuning from Understanding to Manipulation [43.83789393525928]
  InstructVLA is an end-to-end vision-language model that preserves the flexible reasoning of large vision-language models (VLMs) while delivering leading manipulation performance.<n>InstructVLA introduces a novel training paradigm, Vision-Language-Action Instruction Tuning (VLA-IT), which employs multimodal training with mixture-of-experts adaptation.<n>On in-domain SimplerEnv tasks, InstructVLA achieves 30.5% improvement over SpatialVLA.
  arXiv  Detail & Related papers  (2025-07-23T13:57:06Z)
• VILA: On Pre-training for Visual Language Models [74.08039416548209]
  We study the design options for VLM pre-training through step-by-step controllable comparisons. We build VILA, a Visual Language model family that consistently outperforms the state-of-the-art models.
  arXiv  Detail & Related papers  (2023-12-12T18:58:18Z)
• GLIPv2: Unifying Localization and Vision-Language Understanding [161.1770269829139]
  We present GLIPv2, a grounded VL understanding model, that serves both localization tasks and Vision-Language (VL) understanding tasks. GLIPv2 unifies localization pre-training and Vision-Language Pre-training with three pre-training tasks. We show that a single GLIPv2 model achieves near SoTA performance on various localization and understanding tasks.
  arXiv  Detail & Related papers  (2022-06-12T20:31:28Z)
• PEVL: Position-enhanced Pre-training and Prompt Tuning for Vision-language Models [127.17675443137064]
  We introduce PEVL, which enhances the pre-training and prompt tuning of vision-language models with explicit object position modeling. PEVL reformulates discretized object positions and language in a unified language modeling framework. We show that PEVL enables state-of-the-art performance on position-sensitive tasks such as referring expression comprehension and phrase grounding.
  arXiv  Detail & Related papers  (2022-05-23T10:17:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.