Welcome to the M. Mathis Lab of Adaptive Intelligence 👋

We are a team of neuroscientists, computer scientists, and enigneers that come together to tackle one of the largest challanges in science - how does the brain drive adaptive behavior. Namely, our world is always changing: how do our brains adapt? We develop new machine learning methods that enable us to understand the mechanisms underlying adaptive behavior in intelligence systems, aka “adaptive intelligence”.

• From 2017-2023 we were also called the Mathis Lab of Adaptive Motor Control (hence our GitHub org name!) but decided to rename ourselves to better reflect our ML, CV, and systems neuro sides of the lab 🖤. We still really like sensorimotor control though!

Check out our website for more information, and see our open-source code, data, & models below!

Of particular interest:

• DeepLabCut: for user-defined animal and human pose estimation.
• CEBRA: for supervised and unsupervised dimensionality reduction (behavioral data, neural, or both!).
• AmadeusGPT: for using large language models as systems for behavioral analysis.
• LLaVAction: the first multi-model large model for action recognition.
• CellSeg3D: for 3D mesoSPIM segmentation of cell soma.
• DLC2Kinematics for processing DeepLabCut data for kinematic analysis.