English(EN)Does VLA Even Know the Basics? Measuring Commonsense and World Knowledge Retention in Vision-Language-Action Models
新研究增强了用于机器人和视觉推理的VLA模型
作者PulseAugur 编辑部·[43 个来源]·
近期研究探索了增强用于机器人操作和通用视觉推理的视觉-语言-动作（VLA）模型。研究通过域随机化和照片级真实感渲染来研究模拟到现实的泛化能力，并提出诸如 Faithful Warm-Start 等方法，通过在强化学习前确保视觉保真度来提高VLM推理的稳定性。其他工作引入了置信度驱动的测试时强化学习，无需外部奖励即可实现自我改进，以及状态感知分词器，以更好地从离散代码中解码动作。此外，研究还检查了VLA模型中的架构冗余，发现语言骨干对于操作任务高度冗余，并提出了动态技能-工具演化框架，可以在不更新权重的情况下适应冻结的VLM。
AI
arXiv:2607.02490v1 Announce Type: new Abstract: Large vision-language models can reason over multimodal inputs by generating textual chains of thought (CoT). A key capability exhibited in CoT reasoning is self-reflection: revisiting earlier decisions and correcting previous error…
arXiv:2603.18315v2 Announce Type: replace-cross Abstract: Traditional reinforcement learning (RL) methods rely on manually engineered rewards or sparse collision signals, which fail to capture the rich contextual understanding required for safe driving and make unsafe exploration…
arXiv:2607.02092v1 Announce Type: cross Abstract: Flow-matching vision-language-action policies generate robot action chunks through an iterative transport process, creating an opportunity for test-time guidance without retraining the base policy. We study this opportunity in Gui…
arXiv cs.AI
TIER_1English(EN)·Guoyang Xia, Fengfa Li, Hongjin Ji, Lei Ren, Fangxiang Feng, Kun Zhan, Yan Xie·
arXiv:2607.01586v1 Announce Type: cross Abstract: Vision-language-action models (VLAs) have recently advanced robotic manipulation, yet the effects of different robot-data pre-training paradigms remain difficult to compare because existing models often differ in architecture, dat…
Large vision-language models can reason over multimodal inputs by generating textual chains of thought (CoT). A key capability exhibited in CoT reasoning is self-reflection: revisiting earlier decisions and correcting previous errors. However, existing LVLMs often fail to properl…
Flow-matching vision-language-action policies generate robot action chunks through an iterative transport process, creating an opportunity for test-time guidance without retraining the base policy. We study this opportunity in Guided Action Flow, an inference-time framework that …
arXiv cs.AI
TIER_1English(EN)·Amirhosein Chahe, Tyler Naes, Jovin D'sa, Faizan M. Tariq, Sangjae Bae, Lifeng Zhou, David Isele·
arXiv:2607.00283v1 Announce Type: cross Abstract: Autonomous vehicles must safely navigate complex environments where planning-critical agents may be hidden from view. Current approaches often treat all occlusions with uniform conservatism, yielding needlessly defensive driving, …
arXiv:2606.31144v1 Announce Type: cross Abstract: This paper presents an integrated system for the CMU Vision-Language-Action (VLA) Challenge, designed to enable an autonomous agent to perform complex tasks based on natural language instructions. Our framework employs a modular a…
arXiv cs.AI
TIER_1English(EN)·Taozhao Chen, Ian Manchester, Huaming Chen·
arXiv:2606.30686v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) systems, built on pretrained vision-language models (VLMs), have shown rapidly improving performance on robot manipulation benchmarks. These gains are commonly interpreted as evidence that semantic rep…
arXiv:2606.31846v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models offer a promising framework for robotic manipulation by connecting language instructions, visual observations, and continuous control. However, most existing policies remain limited by behavior …
arXiv cs.AI
TIER_1English(EN)·Dongyoon Hwang, Byungkun Lee, Dongjin Kim, Hyojin Jang, Hoiyeong Jin, Jueun Mun, Minho Park, Hojoon Lee, Hyunseung Kim, Jaegul Choo·
arXiv:2606.31329v1 Announce Type: cross Abstract: Hierarchical Vision-Language-Action (VLA) models decouple high-level planning from low-level control to improve generalization in robot manipulation. Recent work in this paradigm uses 2D end-effector trajectories predicted by a Vi…
arXiv:2606.31167v1 Announce Type: cross Abstract: VLA models have emerged as a powerful paradigm for transferring semantic knowledge from web-scale data to physical robotic control. However, current single-frame architectures suffer from intrinsic limitations: temporal myopia tha…
Vision-Language-Action (VLA) models offer a promising framework for robotic manipulation by connecting language instructions, visual observations, and continuous control. However, most existing policies remain limited by behavior cloning or supervised fine-tuning (SFT) from fixed…
arXiv:2606.30185v1 Announce Type: new Abstract: Improving vision-language models (VLMs) on visual reasoning typically requires retraining or hand-designed prompts and tools. We present Dynamo, a training-free framework that adapts a frozen VLM without any weight updates. On a sma…
arXiv:2603.22876v2 Announce Type: replace-cross Abstract: Learning a generalist control policy for robotic manipulation typically relies on large-scale datasets. Given the high cost of real-world data collection, a practical alternative is to generate synthetic data through simul…
arXiv cs.AI
TIER_1English(EN)·Tengyue Jiang, Chunpu Xu, Jiayue Kang, Yao Mu·
arXiv:2606.29984v1 Announce Type: new Abstract: Reinforcement Learning (RL) is an important paradigm for improving the reasoning capabilities of Vision-Language Models (VLMs). However, directly applying RL to rollout multimodal reasoning can lead to instability, due to the exploi…
Improving vision-language models (VLMs) on visual reasoning typically requires retraining or hand-designed prompts and tools. We present Dynamo, a training-free framework that adapts a frozen VLM without any weight updates. On a small labeled training subset, the agent inspects i…
arXiv:2606.27755v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models enable instruction-driven robotic manipulation, but they inherit oversized language backbones from pretrained VLMs whose capacity far exceeds what is needed for short robotic instructions. This …
arXiv:2606.27872v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models have demonstrated strong capabilities in robotic manipulation, but their performance degrades significantly in long-horizon tasks due to cumulative error propagation. This limitation largely ari…
Vision-Language-Action (VLA) models have demonstrated strong capabilities in robotic manipulation, but their performance degrades significantly in long-horizon tasks due to cumulative error propagation. This limitation largely arises from static feature fusion mechanisms that rel…
Vision-Language-Action (VLA) models enable instruction-driven robotic manipulation, but they inherit oversized language backbones from pretrained VLMs whose capacity far exceeds what is needed for short robotic instructions. This raises a basic question: how much of a VLA model i…
arXiv cs.AI
TIER_1English(EN)·Henrik M\"uller, Daniel Kudenko·
arXiv:2606.27180v1 Announce Type: cross Abstract: Sparse rewards are inherently challenging for reinforcement learning agents as they lack intermediate feedback to guide exploration and to correctly attribute the sparse success rewards to relevant parts of the trajectory. Naive r…
arXiv:2510.09976v2 Announce Type: replace Abstract: Vision-Language-Action (VLA) models such as OpenVLA, Octo, and $\pi_0$ have shown strong generalization by leveraging large-scale demonstrations, yet their performance is still fundamentally constrained by the quality and covera…
Research reveals that language backbones in Vision-Language-Action models are highly redundant for robotic manipulation tasks, while vision and action pathways are more critical, suggesting need for deliberate capacity allocation in future architectures.
Sparse rewards are inherently challenging for reinforcement learning agents as they lack intermediate feedback to guide exploration and to correctly attribute the sparse success rewards to relevant parts of the trajectory. Naive reward shaping can induce reward hacking, yielding …
arXiv cs.LG
TIER_1English(EN)·Kejing Wang, Toan Nguyen, Minh Hoang Nguyen, Simon Khan, Flora D. Salim·
arXiv:2606.25800v1 Announce Type: new Abstract: Effective online adaptation of vision-language-action (VLA) models remains challenging, as sparse rewards provide weak supervision for high-dimensional autoregressive action policies. Although self-distillation can in principle prov…
Effective online adaptation of vision-language-action (VLA) models remains challenging, as sparse rewards provide weak supervision for high-dimensional autoregressive action policies. Although self-distillation can in principle provide denser training signals, we find that text-b…
Effective online adaptation of vision-language-action (VLA) models remains challenging, as sparse rewards provide weak supervision for high-dimensional autoregressive action policies. Although self-distillation can in principle provide denser training signals, we find that text-b…
PolicyTrim is a reinforcement learning-based framework that enhances VLA model efficiency by extending reliable action chunk lengths and reducing redundant physical steps through dynamic exploration and redundancy-aware rewards.
Act2Answer protocol evaluates embodied vision-language-action models by having agents answer questions through physical actions, revealing knowledge retention and generalization patterns across different semantic categories.
arXiv cs.CV
TIER_1English(EN)·Yuguang Yang, Canyu Chen, Zhewen Tan, Yizhi Wang, Zichao Feng, Chunyang Liu, Kehua Sheng, Juan Zhang, Linlin Yang, Baochang Zhang, Yan Wang, Bo Zhang, Xianbin Cao·
arXiv:2607.01658v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have emerged as a promising paradigm for end-to-end autonomous driving. However, existing VLAs' training relies heavily on text-centric visual question answering and chain-of-thought reasoning dat…
arXiv cs.CV
TIER_1English(EN)·Mathilde Hochedel, Marc Lalonde·
arXiv:2606.30456v1 Announce Type: cross Abstract: This project investigates whether recent Vision-Language-Action (VLA) models can be transferred from controlled research benchmarks to a real-world robotic platform, specifically a UR5e manipulator, in a reproducible and operation…
arXiv:2606.29384v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have become an important paradigm of embodied AI. However, existing VLA models typically assume well-lit and stable indoor settings, while real-world embodied manipulation may involve degraded RGB…
arXiv:2606.30552v1 Announce Type: cross Abstract: Cross-embodiment transfer in vision-language-action (VLA) models remains challenging because low-level state and action spaces differ fundamentally across robot platforms. We observe that the high-level cognitive process underlyin…
Cross-embodiment transfer in vision-language-action (VLA) models remains challenging because low-level state and action spaces differ fundamentally across robot platforms. We observe that the high-level cognitive process underlying manipulation, including scene perception, object…
This project investigates whether recent Vision-Language-Action (VLA) models can be transferred from controlled research benchmarks to a real-world robotic platform, specifically a UR5e manipulator, in a reproducible and operationally meaningful manner. The work integrates real-r…
arXiv cs.CV
TIER_1English(EN)·Yuan Xu, Yixiang Chen, Kai Wang, Jiabing Yang, Peiyan Li, Qisen Ma, Yan Huang, Liang Wang·
arXiv:2606.26801v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models have shown strong potential for generalizable robotic manipulation. During fine-tuning, however, action supervision applies equally across all timesteps, without structured supervision on which …
Vision-Language-Action (VLA) models have shown strong potential for generalizable robotic manipulation. During fine-tuning, however, action supervision applies equally across all timesteps, without structured supervision on which manipulation stage the robot is in or what the nex…
arXiv:2606.22540v2 Announce Type: replace Abstract: Vision-Language-Action (VLA) models provide a unified paradigm for robotic manipulation, yet their real-world deployment is often bottlenecked by execution efficiency. While existing efforts predominantly focus on compute-centri…
arXiv:2605.11567v3 Announce Type: replace Abstract: Vision-Language-Action (VLA) models predominantly adopt action chunking, i.e., predicting and committing to a short horizon of consecutive low-level actions in a single forward pass, to amortize the inference cost of large-scale…
