Skip to content
@ZJU-REAL

REAL Lab

REAL Lab is a research team at Zhejiang University focusing on Reasoning, Embodied, Agentic and Lifelong Evolving AI.

REAL Lab: Reasoning, Embodied, Agentic, and Lifelong-learning AI at Zhejiang University

Website Google Scholar Hugging Face X RedNote
Total stars   Follow ZJU-REAL

浙江大学 REAL Lab,致力于打造能在真实世界中思考、行动并持续成长的人工智能系统

REAL Lab at Zhejiang University builds AI systems that reason reliably, explore the physical world, act autonomously as agents, and learn continuously throughout their lifetime, bridging the gap between virtual intelligence and physical intelligence.

REAL research map: how Reasoning, Embodied, Agentic, and Lifelong-learning AI connect

Open-source systems

HuggingGPT HuggingGPT   stars
An LLM as controller and the Hugging Face hub as executors: plan, pick expert models, run them, and compose the answer.
HugAgentOS HugAgentOS   stars
A self-evolving AgentOS. Memory, skills, and orchestration improve with every task, and each evolution step is attributable, replayable, and reversible.
Easel Easel   stars
An open-source social-media agent: discover trends, create posts, audio, and video, publish across platforms, and learn what actually works.
ClawGUI ClawGUI   stars
Train, evaluate, and deploy GUI agents in one framework: online RL across parallel emulators and real Android, HarmonyOS, and iOS devices.

More projects →

Research map

R · Reasoning: reason reliably

Models that can reason for as long as a problem requires, spend only the thinking it needs, and know when an answer is right.

Direction Focus Representative work
Long-horizon reasoning Break the context-length ceiling so reasoning can continue as long as the problem demands. InftyThink · InftyThink+
Efficient reasoning Fit the amount of thinking to the difficulty of the problem and stop overthinking. Self-Braking Tuning · LAPO · Mind the Gap
Verifiable reasoning Verifiers and rewards that can be trusted, and models that pause instead of fabricate. VerifyBench · CoVerRL · Pause or Fabricate?

E · Embodied: explore the physical world

Taking intelligence off the screen: understanding 3D space, anticipating how the world changes, and acting through a body.

Direction Focus Representative work
Spatial perception Measure and teach the 3D understanding that 2D pretraining lacks, including viewpoints other than the camera's. ViewSpatial-Bench · SpatialLadder · GPD
Self-evolving spatial intelligence Let geometry and physical interaction supply their own supervision, without human labels. SpatialEvo · Spatial-Interactor
Prediction and world models Reason about what happens next and keep a memory of the world over time. SpaceCast-Bench · WorldMemBench
Embodied action Turn perception into reliable physical action and coordinated embodied behavior. VLA-Corrector · OmniEmbodied

A · Agentic: act autonomously as agents

Agents that plan, use tools, and carry long-horizon tasks through to completion across screens, software, and the open web.

Direction Focus Representative work
GUI grounding Locate exactly where to act on a screen, with dense rewards, test-time learning, and adaptive zoom. GUI-G² · GUI-RCPO · Reliable Negatives · UI-Zoomer
Computer-use agents Long-horizon control of real devices through multi-turn RL, GUI plus command line, and online self-improvement. UI-S1 · EMPO · UI-Copilot · HybridCUA · ComputerSD · Reflect, Revise, Reuse · ClawGUI
Deep research agents Search the open web, synthesize evidence, and carry a project from literature to a reviewed paper. IterSynth · BrowserForge · EviNote-RAG · PaperGym · Polaris
Code agents Code and tests that co-evolve adversarially through reinforcement learning. Code-A1
Creation agents Agents that discover, create, publish across platforms, and learn what actually works. Easel
Tool orchestration LLMs as controllers that plan, call expert models and tools, and compose the result. HuggingGPT · TaskBench
Agent evaluation Interactive, personalized benchmarks and a plug-in harness that evaluates any agent in any environment. KnowU-Bench · ageval

L · Lifelong: learn continuously throughout their lifetime

Deployment is where learning continues: agents turn their own experience into guidance, skills, and better weights.

Direction Focus Representative work
Guided exploration Densify the sparse rewards of long-horizon agentic RL with adaptive hints and milestones. Agent-G² · BEACON
Skill internalization Discover skills in context, then internalize them into the policy across tasks and modalities. SKILL0 · ViSkill · SkillRise
Self-distillation Agentic RL with on-policy self-distillation, and the family of methods built on it. SDAR · AgentOPSD · Pass the Baton · RetireOPD
Post-deployment steering Adjust a deployed model's behavior at inference time without retraining. EasySteer
Self-evolving systems Memory, skills, and orchestration that evolve with use, with every step attributable and reversible. HugAgentOS

If our work helps you, a ⭐ on the repository helps others find it too.

Lab Website  Google Scholar  RedNote  @zjureal  PI Homepage  Join Us

Popular repositories Loading

  1. Easel Easel Public

    An open-source AI agent for social media — discover trends, create content, publish everywhere, and learn what works across Xiaohongshu, Douyin, Zhihu, Bilibili, and more.🎨一个开源的 AI 社交媒体智能体——发现热点趋势、…

    Python 4.7k 788

  2. ClawGUI ClawGUI Public

    Build, Evaluate, and Deploy GUI Agents — online RL training, standardized benchmarks, and real-device deployment in one framework.

    Python 1.4k 59

  3. HugAgentOS HugAgentOS Public

    HugAgentOS: the open-source Agent Harness that runs 24/7

    Python 633 54

  4. Awesome-GUI-Agents Awesome-GUI-Agents Public

    A curated collection of resources, tools, and frameworks for developing GUI Agents.

    469 33

  5. SkillZero SkillZero Public

    [EMNLP 2026] Official code for "SKILL0: In-Context Agentic Reinforcement Learning for Skill Internalization"

    Python 389 21

  6. SDAR SDAR Public

    Official code for "Self-Distilled Agentic Reinforcement Learning"

    Python 381 23

Repositories

Showing 10 of 69 repositories
  • EasySteer Public

    [EMNLP 2026] A Unified Framework for High-Performance and Extensible LLM Steering

    ZJU-REAL/EasySteer's past year of commit activity
    Python 314 Apache-2.0 25 2 0 Updated Nov 16, 2026
  • .github Public

    Organization profile for REAL Lab @ Zhejiang University

    ZJU-REAL/.github's past year of commit activity
    Python 0 1 0 0 Updated Nov 16, 2026
  • HugAgentOS Public

    HugAgentOS: the open-source Agent Harness that runs 24/7

    ZJU-REAL/HugAgentOS's past year of commit activity
    Python 633 54 6 5 Updated Nov 16, 2026
  • ageval Public

    A composable framework for reproducible evaluation of long-running agents.

    ZJU-REAL/ageval's past year of commit activity
    Python 82 MIT 3 4 0 Updated Nov 16, 2026
  • Easel Public

    An open-source AI agent for social media — discover trends, create content, publish everywhere, and learn what works across Xiaohongshu, Douyin, Zhihu, Bilibili, and more.🎨一个开源的 AI 社交媒体智能体——发现热点趋势、创作内容、一键发布至各大平台,并学习分析哪些内容真正有效,覆盖小红书、抖音、知乎、哔哩哔哩等平台。

    ZJU-REAL/Easel's past year of commit activity
    Python 4,722 Apache-2.0 788 5 7 Updated Nov 15, 2026
  • CUA-JEV Public

    Jev for Computer Use

    ZJU-REAL/CUA-JEV's past year of commit activity
    Python 28 3 0 0 Updated Nov 14, 2026
  • decision-agents Public

    Actor, Judge, or Teacher? Understanding Decision Models in Long-Horizon Agents

    ZJU-REAL/decision-agents's past year of commit activity
    Python 47 Apache-2.0 0 0 0 Updated Nov 14, 2026
  • ZJU-REAL/zju-real.github.io's past year of commit activity
    HTML 1 0 0 0 Updated Nov 13, 2026
  • Agent-R2 Public

    Agent-R2: Scaling Agents for Realistic Productivity Work via Real-Source Data Synthesis

    ZJU-REAL/Agent-R2's past year of commit activity
    Python 10 Apache-2.0 0 1 0 Updated Nov 13, 2026
  • KnowU-Bench Public

    Official code for "KnowU-Bench: Towards Interactive, Proactive, and Personalized Mobile Agent Evaluation"

    ZJU-REAL/KnowU-Bench's past year of commit activity
    Python 82 Apache-2.0 2 1 0 Updated Nov 14, 2026

People

This organization has no public members. You must be a member to see who’s a part of this organization.

Top languages

Loading…