← Back to homepage

Caleb Jay Wilson

(goes by Caleb)

Robotics Department, University of Texas at Austin, Austin · github.com/jaywme

Research Interests

Embodied AI, large language models, vision-language-action policies, robot learning, sample-efficient imitation and reinforcement learning, sim-to-real transfer.

Education

University of Texas at Austin, Austin2024 — present
Graduate Student (M.S. → Ph.D.), Robotics — Grounded Embodied Learning (GEL) Lab
Stony Brook University2020 — 2024
B.S. in Computer Science, Honors

Research in Progress

Closed-loop LLM planning for long-horizon manipulation
Ongoing — manuscript in preparation
Data-efficient vision-language-action policies
Ongoing
Calibrated uncertainty for embodied agents
Course project (exploratory)

Experience

Research InternSummer 2025
Embodied AI startup
Graduate Student Instructor2024 — present
University of Texas at Austin — Intro to Robot Learning

Selected Open Source

Skills

Python, PyTorch, JAX, ROS2, MuJoCo, Isaac Sim, C++; distributed training; real-robot data collection.

Awards

This is a personal portfolio CV.