An Observation on Generalization and Unsupervised Compression
Connects Kolmogorov complexity and Solomonoff induction to next-token prediction, arguing that optimal sequence compression forces models to recover shared world structure.
Simons Institute
Faculty & visiting speakers
Talks & colloquia from Simons Institute
2Self-contained seminars, public lectures, and invited talks.
Differential Privacy: Fundamentals to Forefront
Formalizes mathematical privacy guarantees under arbitrary auxiliary information, composition theorems, Laplace/Gaussian noise calibration, and adaptive data analysis.
Study paths featuring Simons Institute
1
6 steps · 6h4 universities
Computability, Zero-Knowledge & Differential Privacy
Automata, cryptographic hardness, interactive proofs, and privacy guarantees
MIT · Ruhr University Bochum · Harvard University +1
6 curated lecturesOpen path →