Tsinghua Institute for Interdisciplinary Information Sciences / 清华大学交叉信息研究院

Overview

The Tsinghua Institute for Interdisciplinary Information Sciences, referred to as 清华叉院 in the source, is the Tsinghua University institute where 徐梦迪 serves as an assistant professor and builds a robot-learning group.

Current Profile

In the bounded interview, the institute is presented as an academic setting that combines strong students, interdisciplinary work, research freedom, and the possibility of integrating hardware, models, data, and real deployment into complete robotic systems. This is Xu’s reason for choosing the institute over a better-resourced industrial route: it supports long-horizon questions and gives students early access to real-machine research and lab building.

Key Characteristics

• Supports interdisciplinary work across robot hardware, machine learning, data, and deployment.
• Offers academic freedom for long-horizon research questions.
• Provides access to strong students and an environment for building a new laboratory.
• Is presented as suitable for system-level robotics rather than algorithm work in isolation.

Evidence

• Institutional choice: “我看到了 Scaling Law 的信号”｜对谈清华叉院助理教授徐梦迪：具身智能、世界模型、真正的泛化 records Xu choosing the institute for its students, interdisciplinary environment, research freedom, and system-building conditions.
• Training culture: “我看到了 Scaling Law 的信号”｜对谈清华叉院助理教授徐梦迪：具身智能、世界模型、真正的泛化 describes open undergraduate participation, tutorial-style group meetings, early real-robot practice, and doctoral students helping build the lab.

Qualifications

The page reflects one faculty member’s reasons for joining and her lab practices, not a complete institutional history, faculty map, admissions guide, or official description of the institute.

What Changed

• Created the institute profile from Xu Mengdi’s account of her institutional choice and laboratory environment.

Relationships

• Tsinghua University / 清华大学 - parent university.
• Xu Mengdi / 徐梦迪 - assistant professor whose lab-building account supplies the current evidence.
• Robot In-Context Learning / 机器人上下文学习 - long-horizon research direction pursued in the represented lab.
• Problem Definition In Research - academic freedom and question selection are central to the institution choice in this source.