Jianxiong Li Get in touch

EMBODIED AI & REINFORCEMENT LEARNING

Jianxiong Li.

李健雄

Co-founder & CEO / BASAL Intelligence

Hi! I study embodied AI and reinforcement learning. My dream is to develop robots that are universally deployable across diverse real-world environments.

I completed my PhD at AIR, Tsinghua University in July 2026, advised by Prof. Xianyuan Zhan and Prof. Ya-Qin Zhang.

Previously, I was a Top Seed at ByteDance Seed Robotics (December 2025–July 2026), working closely with Xiao Ma. I received my bachelor's degree in 2021 from the School of Mechanical Engineering at Xi'an Jiaotong University, where I worked on mechanical design and robotics projects.

Portrait of Jianxiong Li
RESEARCHER & FOUNDERBASAL Intelligence

01 / RESEARCH

Learning to act in the real world.

Three questions that guide my work.

01

Efficient Pretrain

How can we build robotic foundation models efficiently when robotics data are limited?

Explore this work
02

Fast Post-train

How can we rapidly enhance robot performance with a limited budget?

Explore this work
03

RL + X

How can reinforcement learning achieve super-human performance across robots, VLMs, and LLMs?

Explore this work

02 / UPDATES

Recently.

Follow along
Earlier news 11 updates
  • 2026

    Three papers have been accepted to ICLR 2026, including a strong VLA model X-VLA, a stable diffusion RL post-training approch DIPOLE and a sample-efficient offline RL approch TELS.

  • 2025

    🏆Our X-VLA has won 1st place in the AGIBOT World Challenge (Manipulation track) @ IROS 2025.

  • Earlier

    🌟We release X-VLA, a cross-embodiment model that sweeps many benchmarks and achieves strong real-world performance.

  • 2025

    One paper (FlowPlanner) on autonomous driving is accepeted to NeurIPS 2025.

  • 2025

    One paper (LBP) on efficient latent planning is accepted to ICML 2025.

  • 2025

    One paper (UniAct) on cross-embodiment universal actions is accepted to CVPR 2025.

  • 2025

    🌟Diffusion-Planner is selected as oral presentation at ICLR 2025.

  • 2025

    Two papers on fast post-train (PSEC) and autonomous driving (Diffusion-Planner) are accepted to ICLR 2025.

  • 2025

    One paper (Robo-MUTUAL) on embodied representations is accepted to ICRA 2025.

  • 2025

    🌟One paper (RSP) on offline RL is accepted to AAAI 2025 as oral.

  • 2024

    🌟IVM and DecisionNCE are selected as Outstanding Paper at MFM-EAI workshop @ ICML 2024.

03 / PUBLICATIONS

Research, in practice.

Google Scholar

* Equal contribution

28 publications

Publication updates verified against Google Scholar and original papers · September 2026.

04 / EXPERIENCE & EDUCATION

The path so far.

  1. Jul 2026 — PresentCURRENT

    BASAL Intelligence

    Co-founder & CEO

    Co-founded BASAL Intelligence in July 2026.

  2. Dec 2025 — Jul 2026

    ByteDance Seed Robotics

    Top Seed

    Worked closely with Xiao Ma.

  3. Graduated Jul 2026

    Tsinghua University

    PhD · Institute for AI Industry Research (AIR)

    Advised by Prof. Xianyuan Zhan and Prof. Ya-Qin Zhang.

  4. Graduated 2021

    Xi'an Jiaotong University

    Bachelor's degree · School of Mechanical Engineering

    Projects in mechanical design and robotics.

05 / COMMUNITY

Professional service.

Reviewer for ICLR 24–25 (Notable Reviewer), ICML 24–25, NeurIPS 23–25 (Top Reviewer), AAAI 24–25, IJCAI 24, IROS, TMLR, ICLR 2025 Workshop @ World Models, and ICLR 2024 Workshop @ DMLR.

LET'S CONNECT

Join us.

If you're deeply passionate about robot learning and eager to push the limits and advance the field, we'd love to have you join us.

lijianxiong@basalintelligence.com