Unverified50% confidenceEventExact time
通用实验室（Tongying Lab）发布了视觉-语言-动作（Vision-Language-Action）模型
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026
Sources
XAI Releases Grok Build Programming Model, Gemini Spark Opens to Ultra Users
bilibiliinfinite灵感港5/30/2026
Related Claims
Unverified通应实验室推出了Coin VLA视觉语言动作模型，专注于通用具身智能76% similarUnverified该工具的技术基础是多模态大模型与视频理解技术的协同，包括计算机视觉分析画面构图色彩镜头切换、NLP理解字幕配音文案、时序建模捕捉帧间运动和剪辑节奏67% similarUnverifiedMiniMind-V的整体架构设计参考了LLaVA（Large Language and Vision Assistant）的经典范式66% similarUnverifiedBuildBuddy使用多模态大语言模型的视觉理解能力来识别屏幕中的UI控件、文本标签、图标和空间布局关系65% similarVerified多模态大模型通过视觉编码器（Vision Encoder）将图像转化为高维向量表示，再与语言模型的文本理解能力融合64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/50437API
curl https://kongchang.com/api/v1/knowledge/claims/50437MCP
get_claim(id=50437)