Week 18.05.2026 – 24.05.2026

Active matter consists of systems whose constituents continuously consume energy to drive motion far from equilibrium, giving rise to a rich variety of emergent collective behaviours. In this talk, I will discuss two active systems, one synthetic and one biological, that exhibit striking dynamical phenomena. I will first focus on superwalking droplets, where self-propelled droplets coupled to long-lived surface waves display memory-driven dynamics that can be captured by a Lorenz-like dynamical system, leading to increasingly complex behaviour as memory is increased. I will then discuss recent work on the quantitative modelling of collective cell migration during anterior-posterior patterning in the mouse embryo, where intermittent start-stop motion emerges from the rheology and stress relaxation dynamics of the surrounding tissue.

The AdS/CFT correspondence is our most successful working example of the holographic principle, identifying quantum gravity in anti-de Sitter space with a non-gravitational conformal field theory in one dimension lower. In this talk I will discuss how far lessons from AdS/CFT can be pushed towards settings closer to our universe, taking anti-de Sitter’s maximally symmetric cousins, de Sitter and Minkowski space, as a starting point. A key hurdle for holography in these settings is that the physics involves outgoing radiation. I will review results showing that perturbative correlation functions at de Sitter future infinity, and on the celestial sphere of Minkowskispace, can nevertheless be expressed as boundary correlators in Euclidean AdS. This Euclidean perspective provides a concrete bridge between radiative observables and familiar AdS technology, and it helps clarify which structural features of AdS/CFT persist, and which require modification, when one moves beyond the AdS setting.

I will discuss old and new results about the distribution of zeros of modular forms, and relation to Quantum Unique Ergodicity. It is known that a modular form of weight k has about k/12 zeros in the fundamental domain . A classical question in the analytic theory of modular forms is “can we locate the zeros of a distinguished family of modular forms?”. In 1970, F. Rankin and Swinnerton-Dyer proved that the zeros of the Eisenstein series all lie on the circular part of the boundary of the fundamental domain. In the beginning of this century, I discovered that for cuspidal Hecke eigenforms, the picture is very different - the zeros are not localized, and in fact become uniformly distributed in the fundamental domain. Very recently, we have investigated other families of modular forms, such as the Miller basis (ZR 2024, Roei Raveh 2025, Adi Zilka 2026), Poincare series (RA Rankin 1982, Noam Kimmel 2025) and theta functions (Roei Raveh 2026), finding a variety of possible distributions of the zeroes.