Conversations, experiments, and things I’ve made—across neuroscience, machine learning, and robotics.
2024 / Neuroverse
A conversation about what computational models can reveal about the brain, what remains beyond their reach, and where machine learning can take neuroscience next.
With fellow panelists Aaditya Singh, Clémentine Dominé, and Pierre Glaser. Hosted and moderated by Carolina Soares and Clara Lenherr.
2023 / Neuroverse · Episode 53
On the International Brain Laboratory, the value of studying natural behavior, and how generative AI and language models are changing the way we work and think.
In conversation with hosts Carolina Soares and Clara Lenherr.
nanoGPT–Shakespeare learning to talk. These samples compare models trained after approximately 320,000, 640,000, and 960,000 examples, from top to bottom.
My talk on tetrode spike sorting at MIT, alongside my complete tetrode spike-sorting toolbox: data filtering → spike event detection → feature extraction → clustering → evaluation.
A semi-autonomous, GPT-controlled rover robot running on a Raspberry Pi. It combines live video, real-time speech-to-text via Whisper, wireless conversation with GPT, and an interface for GPT to control the motors. Multithreading brings these processes together.
My Neuronauts student mentees’ submission for the 2024 UK Big Bang School Science Competition.