Eric Leonardis, MS, PhD

Postdoctoral Fellow – Talmo Pereira Lab and Eiman Azim Lab – Salk Institute for Biological Studies

PhD – Andrea Chiba Lab – Cognitive Science – UC San Diego

Systems Neuroscience / Reinforcement Learning / Computational Ethology

Scientific Advisor for Film

Eric Leonardis

I research human and animal behavior using methods from systems and computational neuroscience. I am currently working with Talmo Pereira and Eiman Azim on deep reinforcement learning, high-fidelity musculoskeletal simulation, and 3D pose tracking. My research focuses on building interpretable NeuroAI systems by integrating biologically realistic biomechanics with large-scale imitation learning to study the neural computations underlying motor control.

My current postdoc project uses 3D tracking, reinforcement learning, and imitation learning to model rodent forelimb movement. Previously, I worked with Professor Andrea Chiba and Dr. Laleh Quinn on neuroscience experiments involving social behavior toward other conspecifics and interaction with robots. I serve as scientific advisor for the National Academy of Sciences Science and Entertainment Exchange to evaluate the science in film scripts.

New NeurIPS Workshop Paper: High Speeed Imitation Learning for Mouse Forelimb Control.

Our paper “Massively Parallel Imitation Learning of Mouse Forelimb Musculoskeletal Reaching Dynamics” has been accepted to the NeurIPS 2025 – Data on the Brain & Mind: Concrete Applications of AI to Neuroscience and Cognitive Science Workshop, taking place December 7 in San Diego, CA. We utilize MIMIC-MJX which uses JAX and Mujoco-MJX to elicit speeds of more than 1 million steps per second through the physics and RL environment.

Mouse Forelimb Imitation Learning – NeurIPS 2025

"MIMIC-MJX: Neuromechanical Emulation of Animal Behavior" High Speed Imitation Learning From Mocap Data

A new collaborative preprint on our neuromechanical modeling framework, integrating stac-mjx and track-mjx for data-driven control of biomechanical bodies in physics simulation.

MIMIC-MJX Neuromechanical Framework

Comic-Con International — “Evolution of AI in Film and In Reality”

I spoke at Comic-Con International on the “Evolution of AI in Film and In Reality” panel, sharing my work at the intersection of AI, neuroscience, and media. I discussed how modern AI influences storytelling, scientific communication, and ethics. The room was packed, and I’m grateful to Andrea Decker for organizing the event.

Eric Leonardis Comic-Con Panel Photo

I recently gave a talk "Body Horror and the Brain" with the Neurohumanities Network and Harvard Medical School. Check out my lecture below:

I was interviewed in WIRED about body horror and insular cortex. We discussed an HR Giger–inspired survival body-horror video game Scorn and its relationship to embodied cognition and the neuroscience of disgust:

Check out our paper about rodent behavioral and neural responses to robotic agents:

Our nonlinear dynamical systems article about estimating dynamic coupling in the brain is out in Scientific Reports:

I actively educate popular audiences about the role of science and technology in society. I am a recurring guest at KPBS Cinema Junkie Podcast and events with the SD Natural History Museum. I was a panelist at the Horrible Imaginings Film Festival discussing the science behind science fiction.

Check out this video of our rat-robot interaction project in collaboration with Janet Wiles' CIS Lab at UQ: