Seminars

Scattering amplitudes constrain effective field theories from the bottom up through general consistency requirements, including unitarity and causality. Existing bounds mainly use four-point amplitudes. This talk explains how amplitudes with more external points supply additional constraints beyond the four-point analysis, and demonstrates the method through an application to Large-Nc quantum chromodynamics.

Blaschke–Santaló diagrams visualize optimal inequalities connecting specified functionals. This talk introduces their underlying principles and main analytical methods through a detailed theoretical study of the diagram relating perimeter, volume and the first eigenvalue of the Dirichlet Laplacian. It then considers numerical methods for approximating such diagrams. The presentation draws on joint work with Jimmy Lamboley and Eloi Martinet.

Raymundo Arróyave discusses alloy discovery guided by both metallurgical knowledge and Bayesian optimization. The approach combines physical models, high-throughput CALPHAD calculations, microstructural information and experimental feedback to search chemical and processing spaces that are difficult to explore through data alone. Examples involving high-entropy and refractory alloys address multiple objectives and constraints, including feasibility, correlated properties, uncertainty and microstructural sensitivity. The seminar then connects these decision methods to the Autonomous Robotic Metallurgist under development at his university. This platform links modular synthesis, characterization and testing with digital twins, human–robot collaboration and AI agents. The aim is a closed experimental loop in which scientific judgment and physical understanding guide automated execution.

Tatsuya Osaki examines vascular contributions to autism spectrum disorder, focusing on blood–brain barrier integrity and cerebral blood-flow dynamics during early development. Rett syndrome provides a starting point: MeCP2 mutations impair endothelial function, suggesting that vascular cells themselves contribute to disease mechanisms. A three-dimensional microphysiological model built from human induced pluripotent stem cells recreates blood–brain barrier structure and function. Patient-derived endothelial cells form perfusable microvascular networks but show impaired barrier function, a finding independently corroborated in MeCP2-knockout mice. Transcriptomic analyses implicate increased miR-126 expression and disrupted tight junctions; inhibiting miR-126-3p restores barrier function and identifies a potential therapeutic target. The work is extending to hemodynamic changes in other autism subtypes, with two-photon imaging providing in-vivo cross-validation, to identify shared neurovascular mechanisms and candidate interventions.

Vinod Menon examines the brain's default mode network as a foundation for autobiographical memory, self-related processing, and consciousness, and considers how these functions compare with current artificial intelligence systems. The virtual session is part of Stanford's Contemplation by Design Summit.

Vladimir Zakharov presents the tangent-fermion approach to simulating a single chiral fermion on a lattice. Conventional local discretizations that preserve chirality generate unwanted copies of the Dirac cone. Tangent fermions retain one cone and ordinary on-site chiral symmetry by allowing nonlocal hopping. Other elements remain local: the generalized eigenvalue problem, a Euclidean action suitable for quantum Monte Carlo without a sign problem, and an exact matrix-product operator with fixed bond dimension for tensor-network calculations. Applications illustrate Luttinger-liquid behaviour, spontaneous symmetry breaking at a quantum spin Hall insulator’s edge, and symmetric mass generation in the 3-4-5-0 model.

This talk develops a first-principles description of interacting electrons, nuclei and electromagnetic fields by extending density-functional theory to quantum electrodynamics. It introduces electron–photon exchange-correlation functionals and photon many-body dispersion, a generalization of molecular dispersion methods that treats electronic and photonic degrees of freedom together. Applications examine anisotropic coupling, effects beyond single photons and cavity-modified van der Waals forces. The approach aims to retain these strong-coupling effects while remaining efficient enough to simulate large, complex systems inside optical cavities.

Organic semiconductor molecules underpin the light-emitting diodes used in high-performance smartphone and television displays. Richard Friend traces the path from early discoveries to practical OLED technology and examines a wider range of applications, including solar cells, printable electronics and biological interfaces. He then discusses emerging phenomena in these materials and current work on their use as quantum sensors that can be accessed optically.

Producing metal-organic frameworks as nanoparticles could make them easier to process at industrial scale and change properties such as transport. Controlled synthesis and the effects of particle size remain insufficiently understood. Carl K. Brozek presents synthetic methods and mechanistic models for preparing MOF nanoparticles whose electrical conduction, magnetism, optical response and dynamic bonding differ from bulk frameworks. Solution-state spectroscopy and electrochemistry expose size-dependent and interfacial effects that conventional framework materials do not reveal. The work provides approaches for scalable MOF manufacture while raising broader questions about relationships among structure, size and properties. Speaker: Associate Professor Carl K. Brozek, University of Oregon, Department of Chemistry and Biochemistry. Hosted by Jack Zhang and Kevin Kittilstved for Washington State University Chemistry. Monday 19 October 2026, 16:10–17:00 America/Los_Angeles (Pacific Daylight Time). Attend at Fulmer Hall, Room 201, Washington State University, Pullman, Washington, USA, or use the public Zoom Link on the official event page.

Representation learning underlies visual search, retrieval and recognition by encoding gallery images and matching them with queries. Most early work assumes a fixed model, but new training data or more expressive architectures can change the feature representation. Recomputing every gallery vector after an update is expensive for collections containing billions of images and may be impossible when the original images are unavailable because of privacy or storage restrictions. Compatible representation learning aims to update the model while retaining usable existing gallery features. The talk examines the foundational compatibility problem, the role of representation stationarity, and new training techniques that use stationarity to learn compatible feature representations.

Dan Meyer considers the questions generative AI raises for mathematics learning after four years of rapid technological development. He argues that its novelty has been accompanied by educational adoption and institutional change falling short of early expectations, including lower-than-anticipated teacher use. The session explores what the technology can do, what students need and what teaching entails, using these questions to examine AI’s contribution to education.

Vertebrate embryos form repeated body-axis segments, yet the segmentation-clock period does not always match the timing of somite formation. Real-time clock imaging combined with pharmacological and surgical perturbations identifies two timers: a clock that establishes an initial commitment and a subsequent self-organizing hourglass process. Hourglass duration scales with somite size and depends on cytoskeletal regulation. Changing this timer can turn regular zebrafish segmentation into irregular patterns resembling the anterior chicken embryo and alter the number of predetermined compartments and somite sizes. The findings separate commitment from segmentation as interacting modules. QLS–CAMBAM hybrid seminar with Ertuğrul M. Özbudak (Northwestern University), 2026-10-20, 12:00–13:00 Montréal time. Attend in person at 550 Sherbrooke, Room 189, McGill University, or follow the public Zoom link on the official event page. Open to the university community, including faculty, graduate and undergraduate students.

Oden Institute Seminar by Rex Ying (Yale University) on how non-Euclidean geometries, particularly hyperbolic geometry, can enhance foundation models by better capturing hierarchies and symmetries in real-world data, with applications across Transformers, language model training, multimodal systems, and recommender systems.

Ta-Wei Guu examines whether continuous data from wearable devices can reveal aspects of older adults’ depression and dementia that conventional clinical assessments miss. The session covers behavioural patterns, sleep disruption and treatment responses measured outside clinical settings, and considers how repeated observations can track changes in health over time. Guu also presents plans for a technology-enabled longitudinal cohort to study ageing populations over months and years, discussing how the resulting patterns could inform research and clinical practice. The session includes questions and discussion. Guu is an assistant professor and consultant psychiatrist at CMUBH in Taiwan, with an affiliation to King’s College London. Online via Dementia Researcher Communities livestream. Wednesday 21 October 2026, 12:00–12:45 BST (Europe/London; UTC+1). This free event requires a Dementia Researcher Communities account: open the event page, sign in or create an account, and RSVP to access the livestream. Organized by Dementia Researcher, delivered by University College London.

In situ cryo-electron tomography can reveal interactions between pathogens and their hosts inside intact cellular environments. This seminar uses apicomplexan parasites and gram-negative bacteria to explain the workflow from sample preparation through image acquisition and analysis. It covers freezing, cryo-focused ion beam milling, cryo-electron tomography, correlative light and electron microscopy, tomogram reconstruction, segmentation and subtomogram averaging. The presentation discusses which aspects of the host-pathogen interface these approaches can resolve, their advantages and limitations, and when they suit a research question. Familiarity with the organisms or cryo-electron microscopy is helpful but not required.

Jason Lerch presents a translational neuroimaging analysis spanning mouse models and human MRI cohorts to identify biologically informed subgroups across autism and related neurodevelopmental conditions.

Mid-infrared spectroscopy reveals molecular signatures of proteins and lipids, but conventional detection needs specialised equipment. Quantum correlations between differently coloured photons, including entanglement, offer another approach: a mid-infrared photon interacts with the sample while interference enables its information to be measured with visible or near-infrared light. The speakers explain the underlying principle and the progression from quantum spectroscopy to imaging with undetected photons. They discuss their work at the University of Technology Sydney to turn quantum mid-infrared imaging into compact, practical instrumentation for biological research.

Digital health technologies generate continuous real-world observations that complement traditional clinical endpoints. This discussion introduces their use throughout drug research, development and commercialization from a clinical pharmacology perspective. It considers how digital observations differ from conventional data, how clinical pharmacology methods help interpret them, common limitations and pitfalls, and their role in regulatory interactions. Examples examine the conversion of wearable and sensor measurements into validated outcomes and the resulting insights into disease and treatment response. The panel also discusses future directions and the potential advantages of adopting these technologies early. ASCPT members attend free. Nonmember registration remains pending until the US$55 fee is arranged by requesting an invoice from ASCPT. Registration closes 15 minutes before the webinar.

Reconstructing early hominin locomotion helps explain access to food, water, shelter and mates. This talk applies three-dimensional geometric morphometrics to functional variation in the primate postcranial skeleton. High-dimensional sliding semilandmarks, coherent point drift and weighted spherical harmonics capture external shape and internal structure, providing complementary descriptions of skeletal adaptation. Comparisons across living primates inform reconstructions of early hominin locomotor repertoires and the relationship between skeletal morphology and movement. QLS–CAMBAM hybrid seminar with Christine M. Harper (University of Washington), 2026-10-27, 12:00–13:00 Montréal time. Attend in person at 550 Sherbrooke, Room 189, McGill University, or follow the public Zoom link on the official event page. Open to the university community, including faculty, graduate and undergraduate students.

This research webinar examines how neuroscience and settlement research explain refugee trauma, mental health and recovery after resettlement. Belinda Liddell discusses neuroimaging evidence on the effects of visa insecurity, family separation, discrimination and social exclusion on brain function, trauma recovery and psychological wellbeing. The biological consequences of uncertainty are considered alongside implications for interventions and refugee mental health care. Tahereh Ziaian presents an international study of refugee and migrant youth in Australia, Canada and the United States. It identifies factors supporting settlement, social inclusion, wellbeing and lasting integration, with recommendations for policies, services and family support. A moderated expert panel connects these findings to trauma-informed clinical practice, settlement services and policy. The webinar is free.