🌟 Introducing WebAgent Wiki: Your All-in-One Resource! 🌟 Key features: 📚RESEARCH PAPER: 49+ papers categorized by keywords & authors 💻PROJECT: 16+ GitHub projects related to WebAgents 🏢COMPANY:14+ companies contributing to WebAgent development 🔬EXPERT: 300+ experts,

The Microsoft Research Asia StarTrack Scholars Program is a three-month experience designed to foster global collaboration and accelerate frontier research. Applications are due today, December 15. Details on research fields, program description, and application procedures are

The web was built for humans. And honestly, that's fine for you guys. But the next trillion internet users are AI agents, robots, and devices who act on your behalf - booking appointments, filling forms, placing orders, and getting things done - using sites that will never,

Want to use LLMs to build your websites? They may not be as good as shown by prior benchmarks! Introducing🔥FronTalk🔥that evaluates multi-turn multi-modal front-end coding - Users propose new evolving requests in multiple turns, including visual instructions🖼️in the form of

Open models year in review What a year! We're back with an updated open model builder tier list, our top models of the year, and our predictions for 2026. First, the winning models: 1. DeepSeek R1 (@deepseek_ai): Transformed the AI world 2. Qwen 3 Family (@AlibabaGroup): The new

📊 Who’s leading GUI Agent research? I compiled a ranking of top institutions by academic publications in GUI Agents. Key takeaways: 🇨🇳 China shows strong concentration around top universities and big tech labs (Alibaba, Tsinghua, SJTU). 🇺🇸 The U.S. ecosystem is more

AgentBay: A Hybrid Interaction Sandbox for Seamless Human-AI Intervention in Agentic Systems AgentBay is a hybrid interaction sandbox that enhances Human-in-the-Loop (HITL) collaboration in autonomous AI systems by providing secure, isolated environments and an Adaptive

LegalWebAgent: Empowering Access to Justice via LLM-Based Web Agents The LegalWebAgent framework utilizes multimodal large language models to autonomously navigate legal processes and enhance access to justice for ordinary citizens by effectively understanding user queries,

Zoom in, Click out: Unlocking and Evaluating the Potential of Zooming for GUI Grounding This paper introduces ZoomClick, a training-free method that enhances GUI grounding by leveraging zoom properties to improve model accuracy and efficiency, alongside the GUIZoom-Bench

An Index-based Approach for Efficient and Effective Web Content Extraction The paper introduces an Index-based Web Content Extraction method that enhances the efficiency and effectiveness of extracting relevant information from web pages by predicting positional indices instead

MVP: Multiple View Prediction Improves GUI Grounding The paper introduces the Multiple View Prediction (MVP) framework, a training-free approach that enhances GUI grounding stability by aggregating predictions from multiple attention-guided views to mitigate prediction

EcomBench: Towards Holistic Evaluation of Foundation Agents in E-commerce EcomBench is a comprehensive benchmark designed to evaluate the performance of foundation agents in real-world e-commerce environments by incorporating genuine user demands and dynamic market conditions

GAIR: GUI Automation via Information-Joint Reasoning and Group Reflection GAIR is a novel framework that enhances GUI automation by integrating heterogeneous capabilities from multiple Multimodal Large Language Models (MLLMs) through information-joint reasoning and group

🚨#GUIAgent Papers of the Week(12/06～12/12): ◾️GAIR ◾️EcomBench ◾️MVP ◾️IndexLM ◾️ZoomClick ◾️AgentBay ◾️iRAG ◾️LegalWebAgent check it out 👉 https://t.co/UmA4u42yoi

1/3 Very last-minute - but it’s happening. We're hosting a hack day with our friends at CodeRabbit this Saturday. If you’re in SF, come hack with us!

Another DeepSeek moment. This is the world’s first actual smart phone. It’s an engineering prototype of ZTE’s Nubia M153 running ByteDance’s Doubao AI agent fused into Android at the OS level. It has complete control over the phone. It can see the UI, choose/download apps,

Today, we're making Scouts available to everyone! Earlier this year, Scouts was born out of a simple observation — that so many of life's background (or even foreground!) tasks have a recurring flavor, e.g. house hunting, early stages of travel planning, sourcing leads,

Release Day 🎉 Meet OpenApps — a pure-Python, open-source ecosystem for stress-testing UI agents at scale. Runs on a single CPU. Generates thousands of unique UI variations. And it reveals just how fragile today’s SOTA agents are. (Yes, even GPT-4 and Claude struggle.)

fun panel with @jaseweston @ysu_nlp @willccbb @xwang_lk @natashajaques - What agents can/can't solve in 1 yr - 1K+ step tasks - Academia & long-horizon tasks - Continual learning: in-context vs weights - Human-AI co-evolution Claude joined as our first AI panelist! Recording🧵

It was really fun to meet new people and discuss agent environments. Thanks to the workshop organizers for putting together such a great event! Here is the slide deck from the talk:

Thanks everyone for attending our “multi-turn interaction workshop” @mti_neurips ❤️ Hope you had great fun at the workshop and the after-party w/ @PrimeIntellect w/ 🎳🎤🏓 AI people sure know how to party! Let’s keep the “multi-turn interaction” going🤘see you next time 🙌