Thu 16 Apr 11:00: CANCELLED: James Rogers
In person
- Speaker: James Rogers (Kavli Senior Fellow in Exoplanets, IoA, Cambridge)
- Thursday 16 April 2026, 11:00-12:00
- Venue: Thirkill Room, Old Court, Clare College.
- Series: LCLU Coffee Meetings; organiser: Paul B. Rimmer.
Wed 29 Apr 16:00: Descending into the Modular Bootstrap
The modular bootstrap has been a powerful tool for carving out the landscape of allowed two-dimensional conformal field theories (CFTs). In this colloquium, I describe a complementary approach to standard modular bootstrap bounds: using modern machine learning strategies to actively search for CFT spectra that yield a valid torus partition function. Using insights from statistical inference and a custom singular-value-based optimizer, I present evidence for an obstruction to finding CFTs with small central charge and large spectral gaps, and I speculate on what this might imply for the structure of the CFT landscape. Along the way, I reflect on “centaur” approaches to theoretical physics, where human physicists and artificial intelligence collaborate to explore spaces of theories that would be difficult to navigate alone.
- Speaker: Professor Jesse Thaler (MIT / IAIFI)
- Wednesday 29 April 2026, 16:00-17:00
- Venue: MR3.
- Series: Theoretical Physics Colloquium; organiser: Amanda Stagg.
Wed 29 Apr 16:00: Descending into the Modular Bootstrap
The modular bootstrap has been a powerful tool for carving out the landscape of allowed two-dimensional conformal field theories (CFTs). In this colloquium, I describe a complementary approach to standard modular bootstrap bounds: using modern machine learning strategies to actively search for CFT spectra that yield a valid torus partition function. Using insights from statistical inference and a custom singular-value-based optimizer, I present evidence for an obstruction to finding CFTs with small central charge and large spectral gaps, and I speculate on what this might imply for the structure of the CFT landscape. Along the way, I reflect on “centaur” approaches to theoretical physics, where human physicists and artificial intelligence collaborate to explore spaces of theories that would be difficult to navigate alone.
- Speaker: Professor Jesse Thaler (MIT / IAIFI)
- Wednesday 29 April 2026, 16:00-17:00
- Venue: MR3.
- Series: Theoretical Physics Colloquium; organiser: Amanda Stagg.
Wed 29 Apr 16:00: Descending into the Modular Bootstrap
The modular bootstrap has been a powerful tool for carving out the landscape of allowed two-dimensional conformal field theories (CFTs). In this colloquium, I describe a complementary approach to standard modular bootstrap bounds: using modern machine learning strategies to actively search for CFT spectra that yield a valid torus partition function. Using insights from statistical inference and a custom singular-value-based optimizer, I present evidence for an obstruction to finding CFTs with small central charge and large spectral gaps, and I speculate on what this might imply for the structure of the CFT landscape. Along the way, I reflect on “centaur” approaches to theoretical physics, where human physicists and artificial intelligence collaborate to explore spaces of theories that would be difficult to navigate alone.
- Speaker: Professor Jesse Thaler (MIT / IAIFI)
- Wednesday 29 April 2026, 16:00-17:00
- Venue: MR3.
- Series: Theoretical Physics Colloquium; organiser: Amanda Stagg.
Wed 29 Apr 16:00: Descending into the Modular Bootstrap
The modular bootstrap has been a powerful tool for carving out the landscape of allowed two-dimensional conformal field theories (CFTs). In this colloquium, I describe a complementary approach to standard modular bootstrap bounds: using modern machine learning strategies to actively search for CFT spectra that yield a valid torus partition function. Using insights from statistical inference and a custom singular-value-based optimizer, I present evidence for an obstruction to finding CFTs with small central charge and large spectral gaps, and I speculate on what this might imply for the structure of the CFT landscape. Along the way, I reflect on “centaur” approaches to theoretical physics, where human physicists and artificial intelligence collaborate to explore spaces of theories that would be difficult to navigate alone.
- Speaker: Professor Jesse Thaler (MIT / IAIFI)
- Wednesday 29 April 2026, 16:00-17:00
- Venue: MR3.
- Series: Theoretical Physics Colloquium; organiser: Amanda Stagg.
Tue 05 May 13:00: Title to be confirmed
Abstract not available
- Speaker: Azadeh Moradinezhad (LAPTh Annecy)
- Tuesday 05 May 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Mon 15 Jun 13:00: Title to be confirmed
Abstract not available
- Speaker: Speaker to be confirmed
- Monday 15 June 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Mon 08 Jun 13:00: Title to be confirmed
Abstract not available
- Speaker: TBC
- Monday 08 June 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Mon 01 Jun 13:00: Title to be confirmed
Abstract not available
- Speaker: Colin Hill (Columbia)
- Monday 01 June 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Tue 26 May 13:00: Title to be confirmed
Abstract not available
- Speaker: Marta Spinelli (OCA & University of the Western Cape)
- Tuesday 26 May 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Mon 18 May 13:00: Title to be confirmed
Abstract not available
- Speaker: Rachel Gray (Glasgow)
- Monday 18 May 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Tue 12 May 13:00: Title to be confirmed
Abstract not available
- Speaker: Laura Wolz (University of Manchester)
- Tuesday 12 May 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Tue 05 May 13:00: Title to be confirmed
Abstract not available
- Speaker: TBC
- Tuesday 05 May 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Mon 27 Apr 13:00: Title to be confirmed
Abstract not available
- Speaker: Antonio Iovino (NYUAD)
- Monday 27 April 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Thu 19 Mar 16:00: Descending into the Modular Bootstrap
The modular bootstrap has been a powerful tool for carving out the landscape of allowed two-dimensional conformal field theories (CFTs). In this colloquium, I describe a complementary approach to standard modular bootstrap bounds: using modern machine learning strategies to actively search for CFT spectra that yield a valid torus partition function. Using insights from statistical inference and a custom singular-value-based optimizer, I present evidence for an obstruction to finding CFTs with small central charge and large spectral gaps, and I speculate on what this might imply for the structure of the CFT landscape. Along the way, I reflect on “centaur” approaches to theoretical physics, where human physicists and artificial intelligence collaborate to explore spaces of theories that would be difficult to navigate alone.
- Speaker: Professor Jesse Thaler (MIT/IAIFI)
- Thursday 19 March 2026, 16:00-17:00
- Venue: MR3.
- Series: Theoretical Physics Colloquium; organiser: Amanda Stagg.
Wed 18 Mar 14:00: From Protostars to Protoplanets: Towards a Population-level simulation study of gravitationally unstable gas discs
Simulations of collapsing giant molecular clouds show protostars forming with discs that may be massive enough to become gravitationally unstable. This may play a key role both in the formation of the star itself, and in the formation of giant planets through disc fragmentation. We extract a set of young discs (less than 1e5 years) from a global cloud collapse simulation (Bate 2012). We upscale the resolution of these discs using a novel particle splitting method and employ an improved radiative cooling approximation in our runs – both as a requirement to accurately capture fragmenting planetary-mass clumps. This suite of physically-motivated simulations from non-idealised initial conditions should provide a valuable insight into the frequency of disc fragmentation and any resultant protoplanetary population. In this talk I will present the workflow and numerical methods involved in this population study, as well as some key insights from early results
- Speaker: Adam Koval (Edinburgh)
- Wednesday 18 March 2026, 14:00-15:00
- Venue: MR14 DAMTP and online.
- Series: DAMTP Astrophysics Seminars; organiser: Thomas Jannaud.
Mon 16 Mar 16:00: Extreme Plasma Astrophysics
Many spectacular phenomena in the high-energy Universe, such as bright, rapid gamma-ray flares, are powered by complex collective plasma processes around relativistic objects: neutron stars and black holes. While our understanding of such processes has greatly benefitted from traditional (space, solar, laboratory) plasma research, the physical conditions near black holes and neutron stars are so extreme that conventional intuition often fails, and a richer physics framework is required. Extreme astrophysical plasmas are relativistic, interact strongly with radiation, and may be subject to QED (e.g., pair-production) effects. Understanding how this additional physics affects collective plasma processes (waves and instabilities, magnetic reconnection, turbulence, etc.) is the main goal of Extreme Plasma Astrophysics. Exploration of this new exciting frontier is now advancing rapidly, thanks to strong observational motivation, vigorous theoretical efforts, and the advent of novel first-principles relativistic kinetic plasma simulation codes incorporating radiation and QED effects. Laser-plasma experiments will soon also contribute to this revolution. In this talk, I will review the recent progress in this burgeoning new field, focusing on theoretical and computational studies of relativistic radiative magnetic reconnection and turbulence and their astrophysical applications to neutron-star magnetospheres and black-hole coronae and jets. I will also outline key theoretical challenges and future directions of Extreme Plasma Astrophysics.
- Speaker: Prof. Dmitri A. Uzdensky (University of Oxford)
- Monday 16 March 2026, 16:00-17:00
- Venue: MR14 DAMTP and online.
- Series: DAMTP Astrophysics Seminars; organiser: David Hosking.
Fri 20 Mar 13:00: A Nonlinear Endpoint of Charged Horizon Instabilities
Extremal black holes are known to develop a horizon instability in the presence of scalar matter. In this talk, I will discuss how this picture extends to charged perturbations, where the nonlinearities of both Maxwell’s equations and Einstein’s equations lead to qualitatively new behavior. In particular, I will give numerical evidence that dynamical extremal black holes act as universal threshold solutions of the Einstein-Maxwell-scalar field system. These threshold solutions can be understood from the perspective of critical phenomena, with horizon instabilities emerging in the appropriate limit. I will then present numerical results on the existence of arbitrarily large curvatures in the (near-)extremal interior. This curvature may be visible from future null infinity in spacetimes without a black hole region. Based on arXiv:2602.11256.
- Speaker: Zach Gelles (Princeton University)
- Friday 20 March 2026, 13:00-14:00
- Venue: Potter Room / Zoom : https://cam-ac-uk.zoom.us/j/85131795570.
- Series: DAMTP Friday GR Seminar; organiser: Lorenzo Gavassino.
Mon 16 Mar 13:00: More (out of) KiDS galaxies and quasars: going beyond cosmic shear
Kilo-Degree Survey (KiDS), thanks to its depth, excellent imaging and multi-wavelength coverage (jointly with VIKING ), allows us to go beyond cosmic shear analyses. I will present how we extract and employ both low- and high-redshift galaxies and quasars from KiDS, taking advantage of its considerable overlap with various spectroscopic calibration datasets, such as GAMA and now also DESI . In selecting these objects and estimating their photometric redshifts, we use in particular machine-learning approaches including deep learning, and I will discuss our recent results in this matter. I will then overview some of the recent applications of these KiDS photometric galaxies and quasars, based on probes such as galaxy-galaxy lensing and clustering. Time permitting, I will also sketch near-future prospects of extending this kind of studies beyond KiDS with new surveys such as 4MOST and LSST .
- Speaker: Maciej Bilicki (Warsaw, CFT)
- Monday 16 March 2026, 13:00-14:00
- Venue: CMS, Pav. B, CTC Common Room (B1.19) [Potter Room].
- Series: Cosmology Lunch; organiser: Gonzalo Villa.
Thu 12 Mar 16:00: Chance, Necessity, and Cause surrounding the Origin, Evolution, and Nature of Life
(remote)
Biology’s roots in the description of diversity and function have led to a fact-rich empirical understanding of life, which unpacks to statistical reconstructions of Natural History and quite sophisticated mechanistic analyses of structure and function in living systems. The Darwinian pivot, from the previous pinnacle of Humboldtian naturalism and explicitly contra Lamarck, created, from whole cloth, a category of cause acting after the generation of change-events, and parallel to impetus-related notions, in which cause is the origin of change events, that had been the basis of physical theorizing from John of Alexandria through Galileo and then Newton. Darwinian ex-post causation through selection is a close cousin to notions of cause now extensively developed in information theory, which have also back-filled the modern physical theory of matter via thermodynamic ideas.
The foregoing facts, methods, and concepts, as they have been used so far, provide a partial window on the origin of life and its major evolutionary patterns, and an even less-complete framing of the nature of the living state. They have proved to be frustratingly limiting, however, to address questions of the likelihood of life’s emerging as a stage in planetary maturation, of the chance or necessity of any particular living feature, and of the appropriate notion of cause to account for life as a distinctive natural phenomenon. An argument can be made that some of the ideas we need to better address these questions have been developed outside biology, in areas of basic physics, computer science, control engineering, and information theory, where they could be discovered in simpler and more symmetric contexts (and even there, slowly and painfully over more than a century!) and that we can now bring them back into a unified science that includes questions of biological origin, evolution, and nature.
Much of my work has amounted to looking for small, concrete questions about chance, necessity, and cause for different features of life, in which I can give proofs-of-concept for framing the questions in terms of probability, hardness or cost of search, robustness, and related criteria, and use combinatorial tools to propose causal analyses. I have looked at universal core metabolism through this lens for some decades, and recently turned to the origins of macromolecules. Tools include the statistical mechanics of stochastic population processes with stoichiometry (e.g. Chemical Reaction Networks) or generative combinatorial chemistries (e.g. graph grammars). Other interests include basic maths of likelihood in spaces of histories and events, to understand how we should extend thermodynamic concepts, and approaches such as Large Deviation Theory, to the many contexts away from equilibrium where energy conservation may appear but need not be fundamental any longer. Here thermodynamics becomes recognizable as a theory about counting, measuring, and information—albeit discovered and named for the effort to understand heat-generated movement — but not fundamentally about, or limited to, that phenomenal domain.
An overarching question is how we should reason theoretically about complex phenomena in which causation is linked across many scales, in the early stages (where we still are) in which data and essential concepts and insights are still fragmentary and too incomplete to assemble into any coherent worldview.
- Speaker: Eric Smith (Santa Fe Institute)
- Thursday 12 March 2026, 16:00-17:00
- Venue: Battcock, Room F17.
- Series: LCLU Coffee Meetings; organiser: Paul B. Rimmer.