Finite free information inequalities @SimonsInstitute
Finite free information inequalities  @SimonsInstitute
Uploaded October 2026 | Updated November 2026, 3 days ago
Jorge Garza Vargas (MIT)
https://simons.berkeley.edu/talks/jorge-garza-vargas-mit-2026-10-15
Probabilistic Phenomena on Graphs, Groups, and Manifolds

Finite free probability was introduced in the work of Marcus, Spielman, and Srivastava, and was first used to establish the existence of Ramanujan graphs of all sizes.
After discussing some context from von Neumann algebras, information theory, and geometry of polynomials, I will present recent results that establish the finite free analogs of several classical information theory inequalities.
Our results may be interpreted as potential-theoretic inequalities for the zeros of real-rooted polynomials under differential operators that preserve real-rootedness.
This is joint work with Nikhil Srivastava and Zachary Stier.
Finite free information inequalitiesCompressed timestamps and efficient random walks in catalytic spaceRandom unitary circuits with constant spectral gapFrom local discrete curvature notions of graphs to global properties of the underlying spacesProperty (T), sums of square and numerical certificatesA Simple Sub-Polynomial Degree Coboundary ExpanderBipartite matching is in NCDistributional Diffusion ModelsHigh-Dimensional Expansion (HDX) (Part 3)Gap Amplification for Local Hamiltonians with Combinatorial SoundnessLabOS: The AI-XR Co-Scientist That Sees and Works With HumansCoboundary Expansion and Locally Testable Codes (Part 1)
Simons Institute for the Theory of Computing |

Finite free information inequalities

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER