Charles Xu

I am currently a PhD student at UC Berkeley, advised by professor Sergey Levine. I recently interned at Physical Intelligence, where I built a system for real-world online RL with VLA's. Before this, I finished my undergrad at UC Berkeley in EECS in May 2024, where I also conducted research with professor Sergey Levine. I've also had the pleasure of working with Stefan Schaal from Intrinsic. My research interest is learning from real-world experience through on-robot online reinforcement learning for scalable and deployable robot systems.

xuc [at] berkeley [dot] edu  /  GitHub  /  Google Scholar  

Test-Time Gradient Guidance of Flow Policies in Reinforcement Learning

π0.7: a Steerable Model with Emergent Capabilities

RL Token: Bootstrapping Online RL with Vision-Language-Action Models

RLDG: Robotic Generalist Policy Distillation via Reinforcement Learning

Precise and Dexterous Robotic Manipulation via Human-in-the-Loop Reinforcement Learning

Octo: An Open-Source Generalist Robot Policy

SERL: A Software Suite for Sample-Efficient Robotic Reinforcement Learning

Open X-Embodiment: Robotic Learning Datasets and RT-X Models

FMB: A Functional Manipulation Benchmark for Generalizable Robotic Learning

Multi-Stage Cable Routing through Hierarchical Imitation Learning

EECS C106A/206A: Introduction to Robotics

Lab Assistant: Fall 2023

CS 189: Introduction to Machine Learning

Tutor: Spring 2023

Design and source code from Jon Barron's website