Laura Smith

Hello! I'm a final-year PhD student in CS at UC Berkeley advised by Sergey Levine. I spent an amazing summer in 2024 working with the Google DeepMind Robotics team. I'm currently interning at Physical Intelligence. I work on enabling robots to interact with and learn in the real world, so as to acquire human-like abilities. My early PhD was supported in part by the NSF Graduate Research Fellowship, and I am thankful to be supported now by a Google PhD Fellowship.

I am graduating in May 2025 and seeking full-time opportunities in industry research labs. Please connect via email if I might be a good fit for your team!

CV Github G. Scholar Twitter

smithlaura at berkeley dot edu

NewScientist article "Robot dog learns to walk on tough terrain in just 20 minutes"

Berkeley Engineering article "Step by Step"

TechXplore article "A technique that allows legged robots to continuously learn from their environment"

TechXplore article "AVID: a framework to enhance imitation learning in robots"

The following are most representative of my personal work.
For a complete list of publications, please see my Google Scholar profile.

Co-Organizer, CoRL Workshop on Learning to Adapt and Improve in the Real World

2022+

Reviewer, CoRL, RA-L, ICRA, NeurIPS, IROS, ICLR.

Board Member, UC Berkeley Women in EECS

2020+

Co-Organizer, Berkeley AI Research Mentoring Program

2019

Guest Lecture at UC Berkeley CS287: Advanced Robotics (video)

2018-2020

Co-Organizer, Robot Learning Lab Outreach

2018

Guest Lecture at UC Berkeley CS10: The Beauty and Joy of Computing (video)

Laura Smith

STEER: Flexible Robotic Manipulation via Dense Language Grounding

Grow Your Limits: Continuous Improvement with Real-World RL for Robotic Locomotion

Learning and Adapting Agile Locomotion Skills by Transferring Experience

RLPD: Efficient Online Reinforcement Learning with Offline Data

A Walk in the Park: Learning to Walk in 20 Minutes With Model-Free Reinforcement Learning

Legged Robots that Keep on Learning: Fine-Tuning Locomotion Policies in the Real World

PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-Training

AVID: Learning Multi-Stage Tasks via Pixel-Level Translation of Human Videos

SOLAR: Deep Structured Latent Representations for Model-Based Reinforcement Learning