Scholar
Eric J. Michaud
Google Scholar ID: X52GetkAAAAJ
Graduate student, MIT
Citations & Impact
Citations
1,814
H-index
12
i10-index
13
Publications
20
Co-authors
12
list available
Publications
Resume (English only)
Academic Achievements
- Publications: 'The Quantization Model of Neural Scaling' (NeurIPS 2023), 'On the creation of narrow AI: hierarchy and nonlocality of neural network skills' (NeurIPS 2025), and several other papers. Participated in multiple academic talks, including presenting 'The Quantization Model of Neural Scaling' at the MIT Department of Physics workshop and ICLR conference.
Research Experience
- Focused on understanding the internal mechanisms of deep neural networks during his PhD, especially in the area of neural scaling models. Also conducted research on grokking and the structure of neural network representations.
Education
- PhD: Department of Physics at MIT, supervised by Max Tegmark; Undergraduate: Mathematics at UC Berkeley, worked with radio astronomers on SETI, Erik Hoel on deep learning theory, and Adam Gleave at CHAI on AI safety.
Background
- Research Interests: Understanding the internal mechanisms of deep neural networks, particularly how and why they learn. Notable work includes the quantization model of neural scaling, grokking, and the structure of neural network representations.
Miscellany
- Email: eric.michaud99@gmail.com; Twitter: @ericjmichaud_; GitHub and CV available; Google Scholar page provided.
