DAILY INTELLIGENCE · AI APP + VLA

研究日報研究日报

AI App 精選與 VLA 論文評級的每日合輯AI App 精选与 VLA 论文评级的每日合辑。 下方匯總近 7 天的信號總覽，歸檔索引可按領域過濾下方汇总近 7 天的信号总览，归档索引可按领域过滤。

近 7 日匯總近 7 日汇总 07-11 → 07-21
41 AI精選AI精选
191 VLA論文VLA论文
arXiv 週末休刊，VLA 論文將於下個工作日更新 arXiv 周末休刊，VLA 论文将于下个工作日更新
⚡ 3 突破
🔧 68 工具/技術工具/技术
📖 120 背景/觀點背景/观点

⚡ 本週最強信號本周最强信号

AI 今日 · 07-20 VLA 今日 · 07-21
近 3 天內容近 3 天内容 07-19 → 07-21
2026-07-21 22 條条
Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories VLA NeuroCommitSSM: Decision-Centric Shared Autonomy for Safe Assistive Manipulation via EEG-EMG-ET Commit Readiness VLA IMBench: A Benchmark for Intuitive Robotic Manipulation VLA AC-VLA: Robust Out-of-Distribution Action Execution via Compositional Learning VLA Dynamics-Aware Meta-Imitation for Generalization to Unseen Robotic Manipulation VLA Handroid: Bridging Dexterous Hand and Humanoid VLA EgoExoMoCap: Distributed Ego-Exo Human Motion Capture VLA MuxGel: Simultaneous Dual-Modal Visuo-Tactile Sensing via Spatially Multiplexing and Deep Reconstruction VLA RhinoVLA Technical Report VLA NeuralActuator: Neural Actuation Modeling for Robot Dynamics and External Force Perception VLA Jetson-PI: Towards Onboard Real-Time Robot Control via Foresight-Aligned Asynchronous Inference VLA Dichotomous Diffusion Policy Optimization VLA ABot-AgentOS: A General Robotic Agent OS with Lifelong Multi-modal Memory VLA JoyNexus: Service-Oriented Multi-Tenant Post-Training for VLA Models VLA Exo2EgoPose: Leveraging Exocentric Demonstrations for Vision-Language guided Egocentric 3D Hand Pose Forecasting VLA Video = World + Event Stream VLA VTAP Gripper: Synergizing Fingertip Sensing and a Visuo-Tactile Active Palm for Dexterous In-Hand Manipulation VLA Scalable Open-Source Visuotactile Sensor for 6-Axis Contact Wrench Estimation in Tensegrity Robots VLA Difference-Based Relational Learning for Zero-Shot Object-Goal Visual Navigation With Direct Sim-to-Real Transfer VLA Towards Artificial Nerves: Biomimetic Optical-Fiber Tactile Sensing for Robots VLA BayesContact: Uncertain Pose Estimation via Visuo-Tactile Proposals and Simulation-based Inference VLA VTLoc: Learning-based Tactile Contact Localization in Visual Point Clouds VLA
2026-07-20 5 條条
Qwen3.8-Max-Preview 发布：2.4T 参数旗舰模型 AI GitHub Copilot SDK 开源：六语言嵌入 Copilot Agent 工作流 AI code-review-graph：本地优先代码图谱，MCP 精准上下文 AI wigolo：AI 编码 Agent 的本地 web 情报引擎 AI 《深入理解 AI Agent》开源全书：设计原理与工程实践 AI
2026-07-19 6 條条
GPT-5.6 用 prompt 解决凸优化 30 年未解难题 AI xAI 开源 Grok Build 编码 Agent（Rust 实现，支持 ACP 协议） AI Claude Code 控制备用 Mac 完整指南 AI Anthropic 将 Claude Fable 5 正式纳入 Max/Team Premium 套餐 AI HuggingFace 披露 7 月安全事件：遭 AI 自主 Agent 全链路攻击 AI SQLite Query Explainer — Fable 生成的交互式查询计划工具 AI

歸檔索引归档索引

83 期
— 2026 年 7 月 —
VLA
⚡ 1 🔧 4 📖 17 Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories
22 篇
AI
Qwen3.8-Max-Preview 发布：2.4T 参数旗舰模型
5 篇
AI
GPT-5.6 用 prompt 解决凸优化 30 年未解难题
6 篇
AI
GPT-5.6 IQ 首破 130 天才线
6 篇
VLA
⚡ 1 🔧 9 📖 17 Open-AoE: An Open Egocentric Manipulation Dataset and Toolchain for Embodied Learning
27 篇
AI
Kimi K3 发布：Moonshot AI 推出 2.8T 参数模型，即将开源
6 篇
VLA
🔧 15 📖 17 HRIBench: Benchmarking Interaction-Centric Human-Robot Collaboration
32 篇
AI
Inkling: Thinking Machines Lab 发布开源 MoE 模型（975B/41B active）
5 篇
VLA
🔧 9 📖 17 FlowWAM: Optical Flow as a Unified Action Representation for World Action Models
26 篇
AI
DeepSeek 准备最快今年申请 IPO，梁文锋身价飙升至 360 亿美元成 AI 公司新首富
6 篇
VLA
🔧 10 📖 19 EgoSteer: A Full-Stack System Towards Steerable Dexterous Manipulation from Egocentric Videos
29 篇
AI
Apple SpeechAnalyzer API 实测：击败所有 Whisper 模型，速度快 3x
7 篇
VLA
🔧 10 📖 15 FlowDAgger: Human-in-the-Loop Adaptation of Generative Robot Policies in Latent Space
25 篇
AI
geohot: I love LLMs, I hate hype
6 篇
AI
腾讯考虑入股 Manus，不会控股
6 篇
AI
GPT-5.6 全量上线 + ChatGPT Work 发布 + Codex 并入 ChatGPT
5 篇
VLA
⚡ 1 🔧 11 📖 18 LEEVLA: Seeing What Matters in Latent Environment Evolution for Vision-Language-Action
30 篇
AI
GPT-5.6 正式 GA：Luna/Terra/Sol 三模型 + 定价 + 54% Token 效率提升
6 篇
VLA
⚡ 1 🔧 11 📖 18 GeoProp: Grounding Robot State in Vision for Generalist Manipulation
30 篇
VLA
🔧 9 📖 23 Learning 4D Geometric Priors for Inference-Efficient World Action Models
32 篇
AI
Anthropic 发现 Claude 内部「J 空间」：类全局工作台的神经表征
6 篇
VLA
🔧 8 📖 20 Simple-to-Complex Structured Demonstrations for Vision-Language-Action Learning
28 篇
AI
Deep Code 开源 AI 编程助手上线，专为 DeepSeek-V4 适配
5 篇
VLA
📖 1 High-resolution real-time mechanochromic tactile sensors
1 篇
VLA
📖 1 High-resolution real-time mechanochromic tactile sensors
1 篇
AI
GPT-5.6 三大子模型全曝：Sol/Terra/Luna + 速度拨盘，定档 7月7日
7 篇
AI
阿里全面禁用Claude：7月10日起卸载所有Anthropic产品
7 篇
VLA
🔧 11 📖 15 Neuro-Symbolic Safety Guidance for Vision-Language-Action Models via Constrained Flow Matching
26 篇
AI
Anthropic Fable 5 全球上线 + Claude 登陆 Microsoft Foundry
7 篇
VLA
⚡ 1 🔧 11 📖 22 EmbodimentSemantic: A Spatial Scene-Graph Dataset and Benchmark for Vision-Language Models on Embodied Manipulation Trajectories
34 篇
AI
Anthropic 回应 Claude Code 检测中国用户：将回滚隐藏代码
6 篇
VLA
🔧 16 📖 12 Position: Vision-Language-Action Models Cannot Be Verified to Perform Physical Reasoning
28 篇
AI
Claude Sonnet 5 发布：性能接近 Opus 4.8，新 tokenizer 有效涨价 30%
6 篇
VLA
🔧 14 📖 15 Human2Any: Human-to-Robot Transfer via Constraint-Aware Compositional Planning
29 篇
— 2026 年 6 月 —
AI
Meta 因担心模型蒸馏风险 对 Claude 和 Codex 使用施加限制
6 篇
VLA
🔧 10 📖 13 Support-Constrained RL Enables Real-World Policy Improvement without Real-World Experience
23 篇
AI
Semgrep 实测：GLM 5.2 在 IDOR 漏洞检测上击败 Claude Code（39% vs 32% F1）
5 篇
AI
OpenAI 发布 GPT-5.6 预览版：Sol/Terra/Luna 三模型系列
6 篇
VLA
📖 1 Advancing Omnimodal Embodied Agents from Isolated Skills to Everyday Physical Autonomy
1 篇
AI
GPT-5.6 限量预览启动，被迫「一客一审」
6 篇
VLA
⚡ 1 🔧 16 📖 15 E-TTS: A New Embodied Test-Time Scaling Framework for Robotic Manipulation
32 篇
AI
OpenAI 发布首款推理芯片 Jalapeño，九个月完成设计到流片
7 篇
VLA
🔧 15 📖 15 Memory Retrieval in Visuomotor Policies for Long-Horizon Robot Control
30 篇
AI
Claude Tag 发布：@一下Claude即可在Slack派活，Anthropic内部65%代码来自此
6 篇
VLA
⚡ 1 🔧 13 📖 22 Compact Object-Level Representations with Open-Vocabulary Understanding for Indoor Visual Relocalization
36 篇
AI
GPT-5.5-Cyber 发布 + Codex Security 插件 + Codex 日志 Bug
6 篇
VLA
⚡ 1 🔧 15 📖 12 MemoryVAM: Integrating Memory into Video Action Model for Robot Manipulation
28 篇
AI
OpenAI Patch the Planet: AI+专家修复开源漏洞
6 篇
AI
Cloudflare Temporary Accounts for AI agents
6 篇
AI
SpaceX 600亿美元全股票收购Cursor母公司Anysphere
5 篇
VLA
⚡ 1 🔧 13 📖 14 WorkBenchMark: A LEGO-Based Assembly Benchmark with an Assembly-by-Disassembly Baseline for the Smart Manufacturing League
28 篇
VLA
⚡ 1 🔧 22 📖 6 Guava: An Effective and Universal Harness for Embodied Manipulation
29 篇
VLA
🔧 17 📖 11 ACE-Ego-0: Unifying Egocentric Human and Robotic Data for VLA Pretraining
28 篇
VLA
🔧 15 📖 13 $\mu_0$: A Scalable 3D Interaction-Trace World Model
28 篇
VLA
🔧 14 📖 16 $\mu_0$: A Scalable 3D Interaction-Trace World Model
30 篇
VLA
🔧 1 📖 1 MoVerse: Real-Time Video World Modeling with Panoramic Gaussian Scaffold
2 篇
VLA
🔧 10 📖 17 Learning to Assist: Collaborative VLAs for Implicit Human-Robot Collaboration
27 篇
VLA
⚡ 1 🔧 13 📖 14 Making Foresight Actionable: Repurposing Representation Alignment in World Action Models
28 篇
VLA
🔧 12 📖 15 Beyond APIs: Probing the Limits of MLLMs in Physical Tool Use
27 篇
VLA
🔧 14 📖 16 VoLo: A Physical Orchestrator for Open-Vocabulary Long-Horizon Manipulation
30 篇
VLA
🔧 10 📖 18 Robots Need More than VLA and World Models
28 篇
VLA
🔧 2 Let It Be Simple: One-Step Action Generation for Vision-Language-Action Models
2 篇
VLA
🔧 11 📖 13 VISTA: Vision-Grounded and Physics-Validated Adaptation of UMI data for VLA Training
24 篇
VLA
🔧 12 📖 15 See Less, Specify More: Visual Evidence Budgets for Generalizable VLAs
27 篇
VLA
🔧 10 📖 19 World-Task Factorization for Robot Learning
29 篇
VLA
🔧 18 📖 9 ELAN4D: Embodiment-Centric 4D Supervision for Vision-Language-Action Models via Plug-and-Play Adaptation
27 篇
— 2026 年 5 月 —
VLA
⚡ 1 🔧 8 📖 20 GEM: Generative Supervision Helps Embodied Intelligence
29 篇
VLA
🔧 13 📖 13 GEM: Generative Supervision Helps Embodied Intelligence
26 篇
VLA
🔧 9 📖 17 PhyPush: One Push is All You Need for Sensorless Physical Property Estimation with Physics-Guided Transformers
26 篇
VLA
⚡ 1 🔧 12 📖 17 EXPO-FT: Sample-Efficient Reinforcement Learning Finetuning for Vision-Language-Action Models
30 篇
VLA
🔧 10 📖 11 Agentic-VLA: Efficient Online Adaptation for Vision-Language-Action Models
21 篇
VLA
📖 1 Bioinspired ionic thermoreceptors with anisotropic architecture for thermotactile perception in robots
1 篇
VLA
🔧 12 📖 11 Learning Structural Latent Points for Efficient Visual Representations in Robotic Manipulation
23 篇
VLA
⚡ 1 🔧 15 📖 26 Learning Structural Latent Points for Efficient Visual Representations in Robotic Manipulation
42 篇
VLA
🔧 9 📖 25 Rethinking Muon Beyond Pretraining: Spectral Failures and High-Pass Remedies for VLA and RLVR
34 篇
VLA
⚡ 1 🔧 9 📖 18 Key-Gram: Extensible World Knowledge for Embodied Manipulation
28 篇
VLA
🔧 11 📖 20 PhysBrain 1.0 Technical Report
31 篇
VLA
🔧 10 📖 23 SECOND-Grasp: Semantic Contact-guided Dexterous Grasping
33 篇
VLA
🔧 11 📖 17 StereoPolicy: Improving Robotic Manipulation Policies via Stereo Perception
28 篇
VLA
🔧 12 📖 15 StereoPolicy: Improving Robotic Manipulation Policies via Stereo Perception
27 篇
VLA
⚡ 1 🔧 14 📖 15 BioProVLA-Agent: An Affordable, Protocol-Driven, Vision-Enhanced VLA-Enabled Embodied Multi-Agent System with Closed-Loop-Capable Reasoning for Biological Laboratory Manipulation
30 篇
VLA
🔧 10 📖 17 VLA-GSE: Boosting Parameter-Efficient Fine-Tuning in VLA with Generalized and Specialized Experts
27 篇
VLA
⚡ 1 🔧 7 📖 11 From Pixels to Tokens: A Systematic Study of Latent Action Supervision for Vision-Language-Action Models
19 篇