I am interested in mathematical logic, specifically in model theory and continuous logic.
In this talk, we will work within continuous first-order logic, where formulas take truth values in the unit interval. Given an x-type p and a formula f(x,y), we consider the defining map from the y-type space to the unit interval, where the image of a y-type q is the truth value of f(p,q). The properties of such defining maps for both full and local types have been extensively investigated in classical first-order logic, particularly for countable languages and models. Hrushovski, Peterzil, and Pillay established Borel definability of finitely satisfiable types in NIP theories. Later, Hrushovski and Pillay proved strong Borel definability (constructibility) for invariant full types. Simon further studied this question for NIP formulas and invariant local types, and proved another version of Borel definability. In this work, we prove an analogous Borel definability result for continuous logic, following their ideas. Combining this result with the well-known paper of Bourgain, Fremlin, and Talagrand, we obtain a characterization of NIP in terms of Borel definability of approximately realized (finitely satisfiable in continuous logic) types and invariant types, and mention several functional-analytic corollaries.
We introduce the notions of NIP and VC-dimension, and explore their relationship. Using a theorem about VC-dimension, one can derive an analogous result in the setting of pseudofinite structures under the NIP assumption. As a corollary of this in the group context, we obtain a qualitative result for the genericity of formulas with positive pseudofinite counting measure.
Fix an L-theory T and a formula. We will define what it means for a formula to be stable and explore equivalent reformulations, following Shelah’s Unstable Formula Theorem. In particular, we will study the connections between the cardinality of type spaces, n-ladders, n-trees, and the definability of types.
A short exposition of the Connes Embedding Problem and Continuous Logic. Reformulation of CEP in terms of computability is outlined.
An alternative proof of the Stable Graph Regularity Lemma, originally proved by M. Malliaris and S. Shelah. The alternative proof is outlined by C. Terry and J. Wolf.
A short exposition of Chapter 10 of the book The Probabilistic Method written by N. Alon and J. Spencer.
On Vaught's proof that Henkin's construction in the proof of the Completeness Theorem is effective: for a recursive language L, any consistent decidable L-theory has a recursive model.
On the shorter proof provided by Albin J. Jones of one of the partition theorems of P. Erdős and R. Rado.
May 2026. Waco, Texas, US.
May 2025. Las Cruces, New Mexico, US.
November 2024. Chicago, Illinois, US.
May 2024. Ames, Iowa, US.
April 2026. Chicago, Illinois, US.
April 2025. Notre Dame, Indiana, US.
June 2024. Budapest, Hungary.
October 2022. Chicago, Illinois, US.