Max Tegmark - Neural network interpretability: symmetry, geometry and formal verification @IPAMUCLA

Uploaded September 2026 | Updated September 2026, 3 weeks ago

Recorded 01 September 2026. Max Tegmark of the Massachusetts Institute of Technology presents "Neural network interpretability: symmetry, geometry and formal verification" at IPAM's Foundations of Interpretability Workshop.
Abstract: I first survey recent progress in mechanistic interpretability of artificial neural networks, focusing on how symmetry and geometric structure emerge because they help with generalization. I then discuss how recent progress in AI-powered formal verification can help with the ultimate interpretability challenge: enabling neural-network-based AI systems to self-export their machine-learned algorithms and knowledge into formally verified code – much like a human can export the algorithms and knowledge learned by the biological neural network in their brain.
Learn more online at: https://www.ipam.ucla.edu/programs/workshops/foundations-of-interpretability/?tab=overview

Max Tegmark - Neural network interpretability: symmetry, geometry and formal verification

Dimitris Giannakis - Operator-theoretic approach, feature extraction & statistical model of climate

$Artur Izmaylov - Quantum SENiority-based Subspace Expansion (Q-SENSE) and Its Extensions Recorded 18 February 2026. Artur Izmaylov of the University of Toronto presents Quantum SENiority-based Subspace Expansion (Q-SENSE) and Its Extensions at IPAMs Bridging the Gap Between NISQ and FTQC Workshop. Abstract: Quantum SENiority-based Subspace Expansion (Q-SENSE) is a hybrid quantum-classical algorithm that interpolates between the Variational Quantum Eigensolver (VQE) and Configuration Interaction (CI) methods. It constructs Hamiltonian matrix elements on a quantum device and solves the resulting eigenvalue problem classically. Unlike other expansion-based methods — such as Quantum Subspace Expansion (QSE), Quantum Krylov Algorithms, and the Non-Orthogonal Quantum Eigensolver — Q-SENSE introduces seniority operators as artificial symmetries to construct orthogonal basis states. This seniority-symmetry-based approach reduces one of the primary limitations of VQE on near-term quantum hardware — circuit depth — at the cost of measuring additional matrix elements. The artificial symmetries also reduce the number of Hamiltonian terms that must be measured, as only a small fraction of the terms couple basis states in different seniority subspaces. With all these merits, Q-SENSE offers a scalable and resource-efficient route to quantum advantage on near-term quantum devices and in the early fault-tolerant regime. Learn more online at: https://www.ipam.ucla.edu/programs/workshops/bridging-the-gap-between-nisq-and-ftqc/$

Ciaran Ryan-Anderson - Experimental Quantum Error Correction at Quantinuum - IPAM at UCLA

Zhiyan Ding - Ground State Preparation for Near-Term Devices using System-Bath Coupling Dynamics

Miguel Garcia-Garibay - What is being transformed by AI in Higher Education? (Everything)

Ionut-Gabriel Farcas - A tutorial on multi-fidelity methods for fusion energy research, Pt. 1/2

Sebastien Bubeck - A Combinatorics Problem - IPAM at UCLA

Samantha Riesenfeld - Quantum-Classical Algorithms for Biomarker Discovery in Multimodal Cancer Data

Shancong Mou - Derivative-Informed Training of Neural Operators on the Fly - IPAM at UCLA

Chimnay Nirkhe - Making the leap to Quantum PCPs - IPAM at UCLA

Hoifung Poon - Accelerating Medical Discovery - IPAM at UCLA