EVTEvents
AI-for-science events in Cambridge, in one calendar
Talks, workshops and reading groups from across the University, gathered each morning from talks.cam.ac.uk and the Accelerate Programme, with our own programme and the funding deadlines alongside.
Updated · refreshed daily, about 05:00 UTC44 events in the next 60 days
02Agenda
Everything coming up, week by week
Week of Mon 26 Oct 202615 items
Tuesday 27 October
-
Tue 27 Oct13:00Add
Learning task structures in the hippocampus
Xinyu Zhao, Tsinghua University
Abstract
Animals can solve the same problem using distinct strategies, from memorizing specific sensory details to making inferences based on abstract task structure. The mammalian hippocampus has traditionally been viewed as an episodic memory system that encodes the rich sensory details of individual experiences.
-
Tue 27 Oct13:00–14:00Add
Towards a surface-mass balance emulator for Greenland, the European Alps and the Himalayas: lessons learned and plans
Dr. Marco Tedesco; Lamont Research Professor, Marine and Polar Geophysics, Lamont-Doherty Earth Observatory (LDEO), Columbia Climate School Adjunct Scientist, NASA Goddard Institute for Space Studies…
Abstract
Regional climate models (RCMs) such as the Modèle Atmosphérique Régional (MAR) compute surface mass balance (SMB) with physical consistency, but their computational cost and forcing requirements limit large ensembles, long reconstructions and systematic attribution.
-
Tue 27 Oct16:00–17:00Add
Beyond Academia - Financial Times, Chief Data Officer
Kate Sargent, Chief Data Officer, Financial Times
Abstract
The Beyond Academia: AI & Data Science Applications seminar series at the University of Cambridge explores the transformative role of data and artificial intelligence in the world. Designed for postgraduate researchers, early-career academics, and students, the series bridges the gap between high-level research and real-world deployment.
Wednesday 28 October
-
Wed 28 Oct11:45–13:00Add
Cambridge AI in Medicine Seminar - October 2026
Jian Chen and Kai Yi
Abstract
Sign up on Eventbrite: https://medai-oct2026.eventbrite.co.uk Join us for the Cambridge AI in Medicine Seminar Series, hosted by the Cancer Research UK Cambridge Centre and the Department of Radiology at Addenbrooke’s. This series brings together leading experts to explore cutting-edge AI applications in healthcare – from disease diagnosis to drug discovery.
-
Wed 28 Oct13:30Add
Bigger Picture Talk: Can we make cameras as thin as a hair? Building tiny imaging devices with fibres, metasurfaces and AI
George Gordon - Department of Chemical Engineering and Biotechnology, University of Cambridge
Abstract
Conventional cameras and microscopes are built by combining lenses, detectors and illumination systems, with better performance often requiring many large optical components. But what happens when the imaging system needs to be thinner than a millimetre, or even as narrow as a human hair?
-
Wed 28 Oct15:05–15:55Add
Ed Catmull: computers and the movies
Dr Ed Catmull - Co-founder of Pixar
Abstract
Ed Catmull is a computer scientist, co-founder of Pixar Animation Studios, and one of the foundational figures in computer graphics. His research helped establish some of the core techniques of modern 3D computer graphics, and he subsequently led Pixar and Walt Disney Animation Studios through several decades of rapid technological and creative change.
Thursday 29 October
-
Thu 29 Oct15:00–16:00Add
Curvature-Aware Optimization for High-Accuracy Physics-Informed Neural Networks
Khemraj Shukla (Brown University)
Abstract
Efficient and robust optimization is essential for training neural networks, enabling scientific machine learning models to converge rapidly to high accuracy while faithfully capturing complex physical behavior governed by partial differential equations (PDEs).
-
Thu 29 Oct15:00–16:00Add
Toward Reliable Deployment of LLM Agents: From Uncertainty Quantification to Progress Advantage
Sharon Li (University of Wisconsin-Madison)
Abstract
LLM agents are increasingly deployed on long-horizon tasks with tool use, irreversible actions, and unpredictable feedback. Yet we have few principled ways to tell, mid-episode, whether an agent is on track or quietly failing. Most uncertainty quantification (UQ) research still centers on single-turn QA, a poor match for interactive agents.
-
Thu 29 Oct15:00–16:00Add
Digital Twins: From Sensor Design to Real-Time Data Visualisation
Dr. Justas Brazauskas, University of Cambridge
Abstract
Abstract: Digital twins combine real-world data with digital representations of physical systems, but building useful digital twins involves more than simply collecting and displaying data.
-
Thu 29 Oct17:00–18:00Add
Panel session: a conversation on academia in the age of AI
Abstract
The rapid impact and integration of AI into the research environment naturally raises a number of questions and concerns, particularly among PhD students and early career researchers in mathematics and theoretical physics.
Friday 30 October
-
Fri 30 Oct12:00–13:00Add
Talk by Rachel Lawrence · title to be confirmed
Rachel Lawrence (Microsoft Research Cambridge)
-
Fri 30 Oct12:30Add
AI-driven materials discovery
Felix Hanke (CuspAI)
-
Fri 30 Oct13:00–14:00Add
Towards Pixel-Level Earth Observation Understanding with Multimodal Large Language Models
Yujie Zan, University of Cambridge
Abstract
Abstract: Multimodal large language models (MLLMs) are changing how we understand Earth observation imagery, making it possible to interact with satellite images through natural language. However, these models still struggle with tasks that require precise spatial understanding, such as identifying individual objects, mapping land cover, and detecting changes over time.
-
Fri 30 Oct14:00–15:00Add
Learning Memory and Material Dependent Constitutive Laws
Dr George Stepaniants, DAMTP
Abstract
The simulation of multiscale viscoelastic materials poses a significant challenge in computational materials science, requiring expensive numerical solvers that can resolve dynamics of material deformations at the microscopic scale. The theory of homogenization offers an alternative approach to modeling, by locally averaging the strains and stresses of multiscale materials.
-
Fri 30 Oct15:30–16:30Add
Talk by Gašper Beguš · title to be confirmed
Gašper Beguš (University of California, Berkeley)
Week of Mon 2 Nov 20267 items
Tuesday 3 November
-
Tue 3 Nov09:30–13:00Add
Training Workshop: Uncertainty Estimation in AI
Abstract
Neural networks have had success in science and engineering predictive tasks such as computer vision, biology, medicine, and weather. To safely use (make decisions based on) the predictions of neural networks, the predictions need to have reliable uncertainty estimates. Yet, despite their importance, reliable predictive uncertainty estimates are often overlooked in practice.
-
Tue 3 Nov14:00–15:00Add
The Invisible Hand of Stability
Geelon So, ETH Zürich
Abstract
The division of labor enables us to build complex human and machine systems. Today, decentralized multi-agent systems with learned behavior are becoming more prevalent. At the same time, decentralization can come with a costly price of anarchy—agents learn to behave in ways that leave everyone worse off. Can we understand and guard against collectively irrational behavior?
-
Tue 3 Nov14:00–15:00Add
"Operando X-ray correlative imaging and processing of 3D structures inside lithium batteries"
Professor Chun Ann Huang - RSC Marlow Prizewinner 2025 - Imperial College London
Abstract
It is conventionally difficult to measure lithium ion concentrations inside commercially standard battery cell configurations using X-rays because the weak signals of lithium cannot be distinguished from the strong signals of other heavy transition metal elements that are present in batteries.
-
Tue 3 Nov16:00–17:00Add
Out-of-Distribution Generalisation, Predictive Uncertainty, and Continual Learning
Mark van der Wilk, University of Oxford
Abstract
How can we create a learning algorithm that generalises “out of distribution”, i.e. beyond the training data? What can we expect about errors when a model generalises OoD? And what should happen when an error is made? In this talk, we will discuss how OccamLab’s approach to these questions.
Wednesday 4 November
-
Wed 4 Nov14:30–15:30Add
Data-driven acceleration of materials discovery and optimization for energy applications
Professor Silvana Botti, Rohr University Bochum
Abstract
The search for better materials for energy conversion and storage is limited less by our understanding of the underlying physics than by the sheer size of the chemical and structural space to be explored.
Thursday 5 November
-
Thu 5 Nov16:00–17:00Add
Perspectives on Cross-Lingual Consistency in LLMs for Medical Questions
Minh Duc Bui (Johannes Gutenberg University Mainz)
Abstract
Should multilingual LLMs answer medical questions the same way in every language, or adapt to cultural cues? In this talk, we discuss these two perspectives. Multilingual medical benchmarks typically treat cross-lingual variation as error, while cultural adaptation research argues that appropriate answers can legitimately differ.
Friday 6 November
-
Fri 6 Nov12:00–13:00Add
Talk by Lisa Alazraki · title to be confirmed
Lisa Alazraki (Imperial College)
Week of Mon 9 Nov 20268 items
Monday 9 November
-
Mon 9 Nov13:05–13:55Add
Microsoft: Building Future AI Infrastructure: Inside Microsoft Specialized Clouds R&D
Ryan Bailey and Rich Hewitt
Abstract
Join Ryan Bailey and Rich Hewitt, Principal Group Engineering Manager at Microsoft, for a look inside Microsoft Specialised Clouds R&D and how the team is tackling the engineering challenges behind AI agents and infrastructure.
Tuesday 10 November
-
Tue 10 Nov16:00–17:00Add
Beyond Academia - BBC, Head of Applied AI
Nikolay Burlutskiy, Head of Applied AI, BBC
Abstract
The Beyond Academia: AI & Data Science Applications seminar series at the University of Cambridge explores the transformative role of data and artificial intelligence in the world. Designed for postgraduate researchers, early-career academics, and students, the series bridges the gap between high-level research and real-world deployment.
Wednesday 11 November
-
Wed 11 Nov09:30–13:30Add
Training Workshop: Hands on Agentic AI for Science
Abstract
This half-day workshop introduces agentic AI through a mix of short presentations, live demonstrations, and hands-on notebook exercises. Participants will explore how LLM-powered agents can be designed to reason over tasks, use tools, call external APIs, work with structured prompts, and carry out multi-step workflows.
Thursday 12 November
-
Thu 12 Nov10:00–12:00Add
AI Clinic slots - online
Abstract
Have you thought about using AI in your research but aren’t sure how to get started? Or are you already using AI and have run into challenges with implementation? Meet the Accelerate Programme’s team of AI experts to find the support you need.
-
Thu 12 Nov14:00–15:00Add
Talk by Jindřich Libovický · title to be confirmed
Jindřich Libovický (Charles University)
-
Thu 12 Nov15:00–16:00Add
Expander Datacenters: From First Principles to the Cloud
Prof. Michael Schapira, Hebrew University of Jerusalem
Abstract
Abstract: AWS recently adopted an expander-based network topology as the default design for its new datacenters, bringing an idea rooted in graph theory into operation at unprecedented scale. In this talk, we trace the research journey that led to this development.
Friday 13 November
-
Fri 13 Nov12:00–13:00Add
Talk by Anne-Marie Lutgen · title to be confirmed
Anne-Marie Lutgen (University of Luxembourg)
-
Fri 13 Nov16:00–17:00Add
Examining fast hydrological responses to anthropogenic forcing agents using machine-learning climate emulators
Bill Collins, University of East Anglia
Abstract
The response of the climate system to increased greenhouse gases is governed by a combination of fast responses and slow feedbacks. Slow feedbacks are typically activated in response to changes in ocean temperatures on decadal timescales and manifest as changes in climatic state with no recent historical analogues.
Week of Mon 16 Nov 20266 items
Tuesday 17 November
-
Tue 17 Nov14:00–15:00Add
MRC Biostatistics Unit seminar
Pawel Morzywolek, University of Copenhagen
Abstract
Inference on Variable Importance for Treatment Effect Heterogeneity: Shapley Values and Beyond We provide an inferential framework to assess variable importance for heterogeneous treatment effects. This assessment is especially useful in high-risk domains such as medicine, where decision makers hesitate to rely on black-box treatment recommendation algorithms.
Wednesday 18 November
-
Wed 18 Nov14:30–15:30Add
Machine Learning for Materials Discovery: From Screening to Advanced Simulations
Professor Janine George, BAM Berlin
Abstract
The development of advanced materials relies on a continuous feedback loop between computational prediction, experimental synthesis, and characterization. Computational methods support this process in many ways.
-
Wed 18 Nov15:05–15:55Add
Synthesizing Correct Hardware Without Hallucination
Dr Satnam Singh - Harmonic. Visiting academic at the Department of Computer Science and Technology, University of Cambridge
Abstract
Hardware can be formally verified by writing theorems about its desired correct behaviour and then proving these theorems correct against a formal mathematical representation of the circuit. Until now this has been a challenging task requiring rare deep expertise both in using interactive theorem provers and in hardware design and verification.
Thursday 19 November
-
Thu 19 Nov11:00–12:00Add
How Do Language Models Reason? Foundations, Limitations and Neuro-Symbolic Approaches
André Freitas (Idiap Research Institute & University of Manchester)
Abstract
How do language models reason, and what would it take to build AI systems capable of sustained, complex reasoning? This lecture develops the foundations for answering these questions, connecting what we know about the internal workings of language models with methods for evaluating and extending their capabilities.
-
Thu 19 Nov14:00–15:00Add
Lewis Lecture 1: Application of NMR Crystallography to Study Disorder in Mixed-Metal Materials
Professor Sharon Ashbrook, University of St Andrews
Abstract
Inorganic materials play vital roles in our everyday lives; understanding the atomic-scale structure of these solids, and how that affects their physical and chemical properties, is key to the future design of functional materials. Mixed-metal materials offer the opportunity to vary or control properties through changes in the composition and the distribution of atoms.
Friday 20 November
-
Fri 20 Nov12:00–13:00Add
Aligning LLMs to the Human Brain
Mariya Toneva (Max Planck Institute for Software Systems)
Week of Mon 23 Nov 20264 items
Monday 23 November
-
Mon 23 Nov17:30–18:30Add
Do Children and Other Animals Think Like Scientists?
Professor Marta Halina is a Senior Research Fellow at the Leverhulme Centre for the Future of Intelligence
Abstract
For decades, psychologists have characterised young children as little scientists, a comparison made famous by the 1999 book The Scientist in the Crib. A core claim of this perspective is that children (and other animals such as chimpanzees, dogs, and birds) use abstract theories to navigate the world, much like researchers discovering and applying general laws.
Tuesday 24 November
-
Tue 24 Nov10:00–18:00Add
AI for Science Summit 2026
Abstract
AI for Science Summit 2026
-
Tue 24 Nov14:00–15:00Add
Understanding and Protecting Interdependent Privacy in LLMs
Xiao Zhan, University of Cambridge
Abstract
Interdependent privacy arises when one person's use of technology affects the privacy of others. In this talk, I will present an empirical study exploring how and why people disclose others' information in everyday interactions with LLMs, as well as how they perceive and navigate the associated privacy concerns.
Thursday 26 November
-
Thu 26 Nov11:00–12:00Add
Talk by Bálint Gyevnár · title to be confirmed
Bálint Gyevnár (CMU)
Week of Mon 30 Nov 20263 items
Monday 30 November
-
Mon 30 Nov12:30–13:30Add
Thursday 3 December
-
Thu 3 Dec14:00–15:00Add
Talk by Simon Ostermann · title to be confirmed
Simon Ostermann (German Research Centre for Artificial Intelligence)
Friday 4 December
-
Fri 4 Dec16:30–18:00Add
The Psychology of AI: How Artificial Intelligence is Changing Individual, Group, and Organisational Behaviour
Jessica Charter, EY
Abstract
Artificial Intelligence is transforming more than technology. It is reshaping how people think, collaborate, and lead. As AI becomes embedded in everyday work, its influence extends beyond productivity to fundamentally changing human behaviour across individuals, teams, and organisations. This session explores three key questions.
Week of Mon 7 Dec 20262 items
Tuesday 8 December
-
Tue 8 DecAll day
Thursday 10 December
-
Thu 10 Dec11:45–13:00Add
Cambridge AI in Medicine Seminar - December 2026
Shivani Sinha
Abstract
Sign up on Eventbrite: details coming soon Join us for the Cambridge AI in Medicine Seminar Series, hosted by the Cancer Research UK Cambridge Centre and the Department of Radiology at Addenbrooke’s. This series brings together leading experts to explore cutting-edge AI applications in healthcare – from disease diagnosis to drug discovery.
Week of Mon 4 Jan 20271 item
Wednesday 6 January
-
Wed 6 JanAll day
Week of Mon 10 May 20271 item
Thursday 13 May
-
Thu 13 MayAll day
Nothing matches these filters.
03Our programme
Michaelmas 2026
Our first term's programme. Dates are provisional and go out on the mailing list first.
01
Talk · Date to be announced
Founding meeting: AI for science, sorted by what works
What has actually arrived, what has only been demonstrated once, and what is still a promise, followed by the routes in at Cambridge.
02
Journal club · Date to be announced
Journal club: machine-learned interatomic potentials
MACE-MP-0 and Matbench Discovery: what universal potentials get right out of the box, and where they still need fine-tuning.
03
Workshop · Date to be announced
Workshop: protein structure prediction, hands-on
Predict, inspect and question structures on your own laptop, and learn what a confidence score does and does not tell you.
04
Talk · Date to be announced
Talk: learned weather forecasting and neural operators
Why data-driven forecasting reached operations first, and what neural operators can and cannot do for PDEs.
05
Social · Date to be announced
Project-matching evening
Bring a problem or bring a method. We pair domain and machine-learning people for term-long mini-projects.
06
Workshop · Date to be announced
Workshop: Bayesian optimisation in the lab
Design experiments with a Gaussian process in the loop, on a real dataset.
07
Talks · Date to be announced
Lightning talks from members
Five minutes each on what you are building or reading. Propose one when you join.