Physics Department Colloquium

The Large Hadron Collider (LHC) at CERN is the most powerful particle accelerator ever constructed. It enables the study of the fundamental structure of matter by providing proton-proton collisions at the unprecedented energy of 6.8 TeV per beam. It delivers an instantaneous luminosity exceeding 2×1034 cm−2s−1 at its two general-purpose detectors, ATLAS and CMS. During high-intensity operation, the LHC now routinely stores energies of 430 MJ per beam—well beyond its original design specifications.

Host: Mihaly Horanyi

Abstract: TBA

Host: Ana Maria Rey

Abstract: The Schrödinger Cat idea was an early thought experiment intended to point out the weirdness of quantum mechanics. It is a paradigmatic example of the quantum principles of superposition and entanglement. With the vast experimental progress in the last two decades, we can now routinely carry out this experiment in the laboratory.

Abstract: Electrochemistry involves chemical reactions that are driven by the movement of electrons and ions, typically occurring at surfaces or interfaces. A key example is rechargeable batteries, where ions migrate through the liquid electrolyte and electrons flow through the external circuit. The electrochemical reactions take place at the electrode–electrolyte interface where electrode materials receive both ions (Li+, Na+, etc) and electrons during discharging, and release them during charging, enabling the reversible storage of electricity.

Abstract:  

Abstract:   Ultrafast sciences and technologies are founded on the principles of ultrashort-pulse nonlinear optics. Until now, their discrete and bulky nature has hindered the utilization of their vast functionalities for many applications, ranging from sensing to computing and quantum information processing.

Abstract: In this talk, I will take you on a tour of large language models, tracing their evolution from Recurrent Neural Networks (RNNs) to the Transformer architecture. We will explore how Transformers elegantly sidestep the vanishing and exploding gradient issues that plagued RNNs. I will introduce neural scaling laws—empirical relationships reminiscent of scaling behaviors common in physics—that predict how model performance improves with increased computational investment.