Georgia Institute of Technology College of Sciences

Spin glasses are disordered spin systems originated from the desire of understanding the strange magnetic behaviors of certain alloys in physics. As mathematical objects, they are often cited as examples of complex systems and have provided several fascinating structures and conjectures. This talk will be focused on one of the famous mean-field spin glasses, the Sherrington-Kirkpatrick model. We will present results on the conjectured properties of the Parisi measure including its uniqueness and quantitative behaviors. This is based on joint works with A. Auffinger.

Groups, rings, modules, and compact Hausdorff spaces have underlying sets ("forgetting" structure) and admit "free" constructions. Moreover, each type of object is completely characterized by the shadow of this free-forgetful duality cast on the category of sets, and this syntactic encoding provides formulas for direct and inverse limits. After we describe a typical encounter with adjunctions, monads, and their algebras, we introduce a new "homotopy coherent" version of this adjoint duality together with a graphical calculus that is used to define a homotopy coherent algebra in quite general contexts, such as appear in abstract homotopy theory or derived algebraic geometry.

The Hales--Jewett theorem is one of the pillars of Ramsey theory, from which many other results follow. A celebrated result of Shelah from 1988 gives a significantly improved bound for this theorem. A key tool used in his proof, now known as the cube lemma, has become famous in its own right. Hoping to further improve Shelah's result, more than twenty years ago, Graham, Rothschild and Spencer asked whether there exists a polynoimal bound for this lemma. In this talk, we present the answer to their question and discuss numerous connections of the cube lemma with other problems in Ramsey theory. Joint work with David Conlon (Oxford), Jacob Fox (MIT), and Benny Sudakov (ETH Zurich).

For relations {R_1,..., R_k} on a finite set D, the {R_1,...,R_k}-CSP is a computational problem specified as follows: Input: a set of constraints C_1, ..., C_m on variables x_1, ..., x_n, where each constraint C_t is of form R_{i_t}(x_{j_{t,1}}, x_{j_{t,2}}, ...) for some i_t in {1, ..., k} Output: decide whether it is possible to assign values from D to all the variables so that all the constraints are satisfied. The CSP problem is boolean when |D|=2. Schaefer gave a sufficient condition on the relations in a boolean CSP problem guaranteeing its polynomial-time solvability, and proved that all other boolean CSP problems are NP-complete. In the planar variant of the problem, we additionally restrict the inputs only to those whose incidence graph (with vertices C_1, ..., C_m, x_1, ..., x_m and edges joining the constraints with their variables) is planar. It is known that the complexities of the planar and general variants of CSP do not always coincide. For example, let NAE={(0,0,1),(0,1,0),(1,0,0),(1,1,0),(1,0,1),(0,1,1)}). Then {NAE}-CSP is NP-complete, while planar {NAE}-CSP is polynomial-time solvable. We give some partial progress towards showing a characterization of the complexity of planar boolean CSP similar to Schaefer's dichotomy theorem.Joint work with Martin Kupec.