Mikhail Belkin
where else to find them
no profile links on file
what they talk about
where they keep going back
no series more than once — all 2 in the filters
on a stage with
never shared a talk
who talks about the same things
nobody close enough yet
| title | conference | date |
|---|---|---|
| Linear representations of concepts in modern AI models | Smale@95: A Conference in Honor of Steve Smale | 2025-07-21 |
Claim: many human concepts inside LLMs sit on one-dimensional directions recoverable by linear probes, i.e. single index models, and can be found and used to… | ||
| From Classical Statistics to Modern Machine Learning | Frontiers of Deep Learning 2019 | 2019-07-15 |
Argues that modern ML succeeds because many methods interpolate the training set and still generalize: kernels and nets fit MNIST with 60% random labels yet a… | ||
| nothing matches | ||