December 2025 - ITP-quantumseminare

No QIT seminar next week
by Ladina Hausmann 19 Dec '25

19 Dec '25

Hi all, There is no QIT seminar next week. We will resume with the next seminar on February 9th. I wish you happy holidays and a good start to the new year! Best, Ladina

1 0

Tuesday's QIT seminar
by Ladina Hausmann 12 Dec '25

12 Dec '25

Hi all, Next Tuesday, our visitor Frédéric Dupuis will talk about “A brief introduction to the Lean theorem prover”. See below for the abstract. The talk will take place on Tuesday at 11:15 in HIT E41.1 Best, Ladina %%%%%% Title: A brief introduction to the Lean theorem prover Abstract: Lean is an interactive theorem prover: a programming language that can be used to formalize mathematical proofs. The compiler then checks that the proof provided is indeed correct. In this talk, I will give a quick introduction to Lean for interested mathematicians/physicists/computer scientists hoping to one day be able to formalize their research results in this way.

1 0

Tuesday's QIT seminar
by Ladina Hausmann 05 Dec '25

05 Dec '25

Hi all, Next week, Petr Ivashkov will tell us about “Ansatz-free Lindbladian Learning”. See below for the abstract. The talk will take place on Tuesday at 11:15 in HIT E41.1. Best, Ladina %%%%%%%%% Title: Ansatz-free Lindbladian Learning Abstract: A central challenge in quantum device characterization is to reliably reconstruct the dynamics of systems that cannot be simulated classically. The Lindblad master equation provides a principled description of open quantum dynamics. However, learning an unknown Lindbladian from data is a nontrivial identification problem. In this work, we present the first efficient protocol for Lindbladian learning that is capable of identifying the relevant interaction terms without any structural or locality assumptions. Our approach proceeds in two stages: structure learning, which identifies the Pauli terms present in the Hamiltonian and dissipator, and coefficient learning, which estimates the numerical values of their coefficients. For sparse Lindbladians with M unknown coefficients, the sample complexity of our method scales with M and target accuracy epsilon as O(M^4 / epsilon^4). Our protocol requires only the preparation of Pauli eigenstates and measurements in product Pauli bases and does not require an ancilla. This establishes a systematic route to the full characterization of noisy quantum dynamics and provides a foundation for applications ranging from error mitigation to the design of improved quantum devices.

1 0