Seminar by Alan Mishchenko, University of California, Berkeley, 25 Oct 2024, N4-02A-35

Title:  Introducing Open-Source CAD Tool ABC and Its Use to Characterize Overfitting in ML Models

Abstract: The talk presents ABC, an open-source software system for logic synthesis and formal verification developed at UC Berkeley since 2005. On the basic level, ABC is presented in general, what it has to offer for different users, and what are the most important computations and commands. On the advanced level, there is an overview of different ABC packages and the lessons learned while developing them.  The second part of the talk focuses on intrinsic methods to detect overfitting. By intrinsic methods, we mean methods that rely only on the machine learning model and the training data, as opposed to traditional methods that rely on performance on a test set or on bounds from model complexity. We propose a family of intrinsic methods called Counterfactual Simulation (CFS) which analyze the flow of training examples through the model by identifying and perturbing rare patterns. By applying CFS to logic circuits we get a method that has no hyper-parameters and works uniformly across different types of models, such as neural networks, random forests, and lookup tables.

Bio: Alan graduated with M.S. from Moscow Institute of Physics and Technology (Moscow, Russia) in 1993 and received his Ph.D. from Glushkov Institute of Cybernetics (Kiev, Ukraine) in 1997. In 2002, Alan joined the EECS Department at University of California, Berkeley, where he is currently a full researcher. His research is in computationally efficient logic synthesis, formal verification, and machine learning.