<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title>Teach and Grow (TGL)</title>
  <link href="https://tgl.changnie.top/"/>
  <id>https://tgl.changnie.top/</id>
  <updated>2026-08-17T00:00:00Z</updated>
  <author><name>Chang Nie</name></author>
  <entry>
    <title>Teach and Grow — project page</title>
    <link href="https://tgl.changnie.top/"/>
    <id>https://tgl.changnie.top/</id>
    <updated>2026-08-17T00:00:00Z</updated>
    <summary>Teach and Grow (TGL) is a training-free robot-learning architecture: an AI agent turns sparse teaching into reusable Skill Blocks, acquiring new manipulation tasks with fixed pretrained weights and physical feedback. 99.9% mean success across four LIBERO suites and 92.4% across seven LIBERO-Plus perturbation categories.</summary>
  </entry>
  <entry>
    <title>Teach and Grow — 项目主页</title>
    <link href="https://tgl.changnie.top/zh/"/>
    <id>https://tgl.changnie.top/zh/</id>
    <updated>2026-08-17T00:00:00Z</updated>
    <summary>Teach and Grow（TGL）是一种免训练的机器人学习架构：AI 智能体把少量示教转化为可复用的 Skill Block，在预训练权重固定的条件下，通过物理反馈完成新任务。在四个 LIBERO 套件上平均成功率 99.9%，在七个 LIBERO-Plus 扰动类别上 92.4%。</summary>
  </entry>
</feed>
