让视觉语言动作模型一步生成动作，效果更好且更简单。
Let It Be Simple: One-Step Action Generation for Vision-Language-Action Models
- 从观测到动作的单步生成比文本到图像更高效
- 高噪声训练下在LIBERO数据集上达到95.6%成功率
- 适合追求高效、简洁动作生成的机器人研究者
从稀疏文本生成多样化图像困难，而从丰富观测生成紧凑动作则更容易。从条件-目标视角看，视觉语言动作（VLA）模型应与图像到文本任务对齐，而非文本到图像。我们通过标准流匹配的不可约速度损失 $R_v(t,c)$ 形式化这一观点，并在受控的8模式模拟实验和MNIST图像到文本任务中验证。随后发现，高噪声训练可显著提升VLA在标准LIBERO上的单步解码性能，在LIBERO-Long上达到95.6%准确率，且在LIBERO-Plus、LIBERO-Pro及真实机器人任务中保持竞争力；而削弱条件或扩展预测时域的消融实验则可预测地消除单步优势。结果表明，单步动作生成是否有效取决于条件-目标结构，而非特殊训练策略。
原文摘要 · Abstract (English)
Generating diverse images from sparse text is hard; generating compact actions from rich observations is easier. From the condition-target view, Vision-Language-Action (VLA) thus aligns with image-to-text, not text-to-image. We formalize this view through the irreducible velocity loss $R_v(t,c)$ of standard flow matching and validate it with a controlled 8-mode toy experiment and image-to-text MNIST task. We then show that high-noise training boosts one-step VLA decoding on standard LIBERO, achieving 95.6% on LIBERO-Long, and remains competitive across LIBERO-Plus, LIBERO-Pro, and real-world robot tasks, while ablations that weaken the condition or expand the horizon predictably erase the one-step gain. These results suggest that whether one-step action generation works in VLA depends not on specialized training, but on the condition-target structure.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。