For information about attending a Salisbury Cafe Scientifique event, see the Attending an Event section; there is also more general information in the Frequently Asked Questions section and help on making the most of this calendar in the Calendar Help section. If you fancy a night of science outside but close to Salisbury, there is also this filtered list of nearby events.

Computational Neuroscience Journal Club – Yul Kang and Wayne Soo @ Online on Zoom
Oct 20 @ 3:00 pm – 4:30 pm

Please join us for our fortnightly journal club online via zoom where two presenters will jointly present a topic together. The zoom link will be sent nearer the time.

The next topic is ‘recurrent neural network (RNN) models of spatial navigation’.

RNNs are suited for the study of circuits involved in spatial navigation because (1) unlike feedforward networks, they are capable of maintaining an internal state (e.g., current location of the agent/animal in the arena) and updating it, which is necessary for navigation, and (2) the brain regions involved in spatial navigation (hippocampal-entorhinal system) are known to have recurrent connectivity that is important for maintaining their spatial representation.

In Part 1, we will look at some early and straightforward approaches that directly use RNN models. Kanitscheider et al. trained their network to perform simultaneous location and mapping. Banino et al. used an RNN to perform path integration, and investigated the efficiency of its resultant grid-like representations. Cueva et al. tackled a similar path integration task with their own RNN model, which gave rise to various spatial-selective units such as grid and band cells.

In Part 2, we will cover recent proposals that push the boundary of the field by studying unsupervised training or incorporating more biological structure into the model. Recanatesi et al. trained their network without supervision using predictive learning, and offered an explanation why predictive learning gives rise to low-dimensional representation of latent variables. Evans et al. incorporated known hippocampal-entorhinal structure into their model and explained the observed pattern of the hippocampal-entorhinal activity that deviates from what would be expected from a simple rule of the physical space.


Kanitscheider, I., Fiete, I. Training recurrent networks to generate hypotheses about how the brain solves hard navigation problems. NeurIPS (2017).

Banino, A., Barry, C., Uria, B., Blundell, C., Lillicrap, T., Mirowski, P., Pritzel, A., Chadwick, M., Degris, T., Modayil, J., Wayne, G., Soyer, H., Viola, F., Zhang, B., Goroshin, R., Rabinowitz, N., Pascanu, R., Beattie, C., Petersen, S., Sadik, A., Gaffney, S., King, H., Kavukcuoglu, K., Hassabis, D., Hadsell, R., Kumaran, D. (2018). Vector-based navigation using grid-like representations in artificial agents. Nature

Cueva, C., Wang, P., Chin, M., Wei, X. Emergence of functional and structural properties of the head direction system by optimization of recurrent neural networks. ICLR Spotlight (2020).

Recanatesi, S., Farrell, M., Lajoie, G., Deneve, S., Rigotti, M., Shea-Brown, E. Predictive learning extracts latent space representations from sensory observations. bioRxiv (2019).

Evans, T., Burgess, N. (2020). Replay as structural inference in the hippocampal-entorhinal system. bioRxiv

NeuroCOVID: Epidemiology, biomarkers, and pathophysiology – Professor David Menon @ Register on Zoom
Oct 20 @ 4:00 pm – 5:00 pm

Abstract to follow

“Click here to register in advance for this meeting”:

Or copy this link into your browser window:

Quantum Computing and High-energy physics – Heather Gray (UC Berkeley/LBNL) @
Oct 20 @ 4:00 pm – 5:00 pm

High-energy physics is facing a daunting computing challenge with the large datasets expected from the HL-LHC and other future colliders. Currently available quantum computers range from D-Wave commercial quantum annealers to gate-based quantum processor prototypes. I will discuss recent developments in quantum computing and provide an overview of recent work that may lead to solutions for problems in high-energy physics.

The first Egyptian society – Anna Marie Roos (University of Lincoln) @ Zoom
Oct 20 @ 5:00 pm – 6:30 pm

Why is sleep good for you? – Professor William Wisden FMedSci, Department of Life Sciences, Imperial College London @ Google Meets
Oct 20 @ 6:00 pm – 7:30 pm

*Register your interest here:*


We have all felt the effects of a poor night’s sleep. If we don’t sleep well, we feel generally groggy. And if we don’t sleep at all, our body takes over and forces us to go to sleep, at least for a while. It seems that regular sleep is needed to maintain the body’s health. But researchers still do not understand why we need to sleep. In this talk, I will outline the current research findings about how sleep may help the brain clear the build-up of waste toxins, stimulate memory, stabilize mood and boost the immune system. I will present our research on the brain’s wiring that controls the drive to sleep.

*Speaker profile:*

Prof William Wisden studied Natural Sciences (Zoology) at the University of Cambridge, and then did his PhD with Prof Stephen Hunt at the MRC Molecular Neurobiology Unit, Cambridge, followed by a period as postdoc in Prof Peter Seeburg’s lab at the University of Heidelberg, Germany. He became Professor at Imperial College London in 2009 and is now the Chair of Molecular Neuroscience there. Prof Wisden has worked extensively on neurotransmitters and the neural circuitry of memory. Most recently, he became interested in sleep. In collaboration with Prof Nick Franks (also at Imperial), he used mouse genetics to investigate how inhibition regulates the sleep-wake circuitry and the actions of sedative drugs. Prof Wisden is a Fellow of the Academy of Medical Sciences.

Politics and Policy – Hannah Clapham (National University of Singapore)
Oct 21 @ 9:30 am – 11:30 am
Systematic improvement of molecular excited state calculations by inclusion of nuclear quantum motion – Antonios Alvertis @ Zoom
Oct 21 @ 11:30 am – 12:30 pm

Many theoretical studies of excited state molecules aim to provide accurate solutions to the electronic Schrödinger equation in order to produce energies that can be compared to experiment. However, nuclear quantum motion, which is usually ignored, can also affect exciton energies, as we showed in a recent study [Alvertis et al. Physical Review B, 102, 081122(R) (2020)]. Here we provide an intuitive picture for the effect of nuclear quantum motion on exciton energies and find that zero-point nuclear quantum fluctuations can significantly affect the energies of low-lying excited states. We compute the vibration-induced corrections to exciton energies on a large set of diverse molecules by combining TDDFT with Monte Carlo sampling techniques based on finite difference methods. We show that incorporating nuclear zero-point energy effects can lead to corrections of up to 1.1 eV on computed exciton energies. We compare our results with a benchmark set of molecules in the literature [Schreiber et al. Journal of Chemical Physics, 128, 134110 (2008)] finding that the correction to excited state energies by incorporating nuclear quantum motion, and without any adjustable parameters, leads to vastly improved agreement with experimental results, while maintaining a low computational cost. We therefore establish nuclear quantum motion as a critical factor towards the accurate calculation of exciton energies.

Join Zoom Meeting:
* Meeting ID: 857 0241 5099
* Passcode: 501623

Science and speculation – Adrian Currie (University of Exeter) @ Teams
Oct 21 @ 1:00 pm – 2:30 pm

Despite wide recognition that speculation is critical for successful science, philosophers of science have attended little to it. When they have, speculation has been characterized in narrowly epistemic terms: a hypothesis is speculative due to its (lack of) evidential support. These accounts provide little guidance to what makes speculation productive or egregious, and how to foster the former while avoiding the latter. I examine how scientists discuss speculation and identify various functions speculations play. On this basis, I provide an account which starts with the epistemic function of speculation. This analysis grounds a richer discussion of when speculation is egregious and when it is productive, based in both fine-grained analysis of the speculation’s purpose, and what I call the ‘epistemic situation’ scientists face.

Emerging Infections – C. Jessica Metcalf (Princeton University)
Oct 21 @ 2:00 pm – 4:00 pm
Seasonal prediction and predictability of regional Antarctic sea ice – Mitch Bushuk, Geophysical Fluid Dynamics Laboratory
Oct 21 @ 2:00 pm – 3:00 pm

Compared to the Arctic, seasonal predictions of Antarctic sea ice have received relatively little attention. In this talk, I will use three coupled dynamical prediction systems developed at the Geophysical Fluid Dynamics Laboratory to assess the seasonal prediction skill and predictability of Antarctic sea ice. These systems, based on the FLOR, SPEAR-lo, and SPEAR-med dynamical models, differ in their coupled model components, initialization techniques, atmospheric resolution, and model biases. This allows for an investigation of these factors in determining Antarctic sea ice prediction skill. Using suites of retrospective initialized seasonal predictions spanning 1992-2018, we find that each system is capable of skillfully predicting regional Antarctic sea ice extent (SIE) with skill that generically exceeds that of a persistence forecast. Winter SIE is skillfully predicted up to 11 months in advance in the Weddell, Amundsen and Bellingshausen, Indian, and West Pacific sectors, whereas winter skill is notably lower in the Ross sector. Zonally advected upper ocean heat content anomalies are found to provide the crucial source of prediction skill for the winter sea ice edge position. The SPEAR systems are notably more skillful than FLOR for summer sea ice predictions, owing to improvements in sea ice concentration and sea ice thickness initialization. Summer Weddell SIE can be skillfully predicted up to 8 months in advance in SPEAR-med, due to the persistence and drift of initialized sea ice thickness anomalies from the previous winter. Overall, these results suggest a promising potential for providing operational regional Antarctic sea ice predictions on seasonal timescales.

Cloud-mounted Molecular Experiments during COVID times – Dr David Glowacki, University of Bristol @ Zoom Meeting ID: 986 1567 0048 Passcode: 266539
Oct 21 @ 2:30 pm – 3:30 pm

In this talk, I will describe some recent attempts to carry out molecular science studies during COVID times. Specifically, I will describe two projects:
(1) Crowd-sourced attempts to search the space of ML strategies and develop algorithms for predicting atomic-pairwise nuclear magnetic resonance (NMR) properties in molecules. Using an open-source dataset, we worked with Kaggle to design and host a 3-month competition which received 47,800 ML model predictions from 2,700 teams in 84 countries. Within 3 weeks, the Kaggle community produced models with comparable accuracy to our best previously published ‘in-house’ efforts. A meta-ensemble model constructed as a linear combination of the top predictions has a prediction accuracy which exceeds that of any individual model, 7-19x better than our previous state-of-the-art. [1]
(2) Efforts to develop Narupa, [2] a flexible, open-source, cloud-mounted, multi-person VR software framework which enables groups of researchers distributed across the world to simultaneously cohabit real-time simulation environments and interactively build, inspect, visualize, and manipulate the dynamics of complex molecular structures with atomic-level precision. [3,4] I will outline a range of application domains where we are using Narupa to obtain microscopic insight into 3D dynamical concepts and enable effective research and communication, including protein-ligand binding, [5] and machine learning potential energy surfaces. [6]

1. L. A. Bratholm et al., “A community powered search of machine learning strategy space to find NMR property prediction models,” arxiv. 2008.05994
2. M. O’Connor et al., An open-source multi-person virtual reality framework for interactive molecular dynamics: from quantum chemistry to drug binding, J. Chem Phys 150, 224703, 2019.
3. M. O’Connor et al., Sampling molecular conformations and dynamics in a multiuser virtual reality framework. Science Advances, 2018, 4 (6).
5. H. M. Deeks et al., “Sampling protein-ligand binding pathways to recover crystallographic binding poses using interactive molecular dynamics in virtual reality”, arXiv:1908.07395, 2019
6. S. Amabilino et al., Training neural nets to learn reactive potential energy surfaces using interactive quantum chemistry in virtual reality. J Phys Chem A, 2019, 123, 20, 4486, 2019

How is Human Social Cognition Special? – Dr Lasana Harris (UCL) @ via zoom
Oct 21 @ 4:00 pm – 5:00 pm

Humans cannot claim to be the only social animal. But is there something special about human social cognition? Here I argue that it distinguishes itself by its unique flexibility: In scope, as it can be extended beyond fellow living humans and withheld from them; in content, as it is revised over time, and in certainty; in process, as it follows Bayesian inferences based on probability. Further, I posit that both the hyper-sociality of humans and some of its distinctive anthropological manifestations derive from this flexibility. This theoretical approach provides an alternative to content-based accounts of superior and distinctive human social cognition, and challenges the static ‘social-wiring’ hypothesis. Here, I present social neuroscience data consistent with this alternative hypothesis.

Dr Harris will discuss how this perspective can inform our understanding of racial prejudice in society and in research.

Dr Lasana Harris is a Senior Lecturer in Experimental Psychology at University College London.

This talk is co-hosted with the Zangwill Club.

Zoom link to follow

How is Human Social Cognition Special? – Professor Lasana Harris, Associate Professor, Experimental Psychology, University College London @ Zoom meeting
Oct 21 @ 4:00 pm – 5:30 pm

Humans cannot claim to be the only social animal. But is there something special about human social cognition? Here we argue that it distinguishes itself by its unique flexibility: In scope, as it can be extended beyond fellow living humans and withheld from them; in content, as it is revised over time, and in certainty; in process, as it follows bayesian inferences based on probability. Further, we posit that both the hyper-sociality of humans and some of its distinctive anthropological manifestations derive from this flexibility. This theoretical approach provides an alternative to content-based accounts of superior and distinctive human social cognition, and challenges the static ‘social-wiring’ hypothesis. Here, I present social neuroscience data consistent with this alternative hypothesis.


Dr. Harris completed his undergraduate education at Howard University, USA, and received post-graduate training at Princeton University, USA. He has held positions at New York University, USA, Duke University, USA, and Leiden University, the Netherlands, before coming to UCL. Dr. Harris’ research explores the neural correlates of person perception, prejudice, dehumanization, anthropomorphism, social learning, social emotions, empathy, and punishment. This research addresses questions such as: How do we see people as less than human, and non-human objects as human beings? How do we modulate affective responses to people? How do we decide right from wrong?

Space photovoltaics challenges: fundamental to applied – Dr Louise Hirst, Semiconductor Physics, Cavendish Laboratory & Department of Materials Science, Cambridge
Oct 21 @ 4:00 pm – 5:00 pm

Abstract not available

Finding Hidden Order with your Cell Phone – Prof. Paul Chaikin, New York University @ Online:
Oct 21 @ 7:00 pm – 8:30 pm

“We introduce a simple idea enabling the quantification of organization in non-equilibrium and equilibrium systems, even when the form of order is unknown. The length of a losslessly compressed data file is a direct measure of its information content. Just compare the size of a picture file before and after you push the “compress” key on your cell phone or computer. We use lossless data compression to study several equilibrium and out-of-equilibrium systems, artificial colloidal swimmers, bacteria, and other active particles, to explore ordering, dynamical phase transitions, and critical behavior.”

Interventions – Mirjam Kretzschmar (Universitair Medisch Centrum Utrecht; Rijksinstituut voor Volksgezondheid en Milieu (RIVM))
Oct 22 @ 9:30 am – 11:30 am
Tumour structure and nomenclature – Dr Paul Edwards, Emeritus Reader in Cancer Biology @ Online
Oct 22 @ 9:30 am – 10:30 am

UPDATE: This lecture will be online only. Please use the following link to apply to be admitted to the talk.

Values-based Practice in Psychiatry – Dr Robert Dudas, Cambridgeshire and Peterborough Foundation Trust & Department of Psychiatry @ Webinar (via Zoom online)
Oct 22 @ 12:30 pm – 1:30 pm

Abstract not available

An Optimal Control Approach to Covid-19 – Prof. Jan Maciejowski, University of Cambridge @ Online (Zoom)
Oct 22 @ 2:00 pm – 3:00 pm

Most countries have controlled the spread of the Covid-19 pandemic by reducing social interaction. This has drastically reduced their economic activity, and a legitimate question is “Is the cure worse than the disease?” In this presentation we show a methodology for answering this and related questions, and for assessing the appropriate policy. The problem is posed as a classical optimal control problem and solved numerically under various assumptions. We assume that no vaccine is available for one year, so that the control input is the severity of social regulation. The costs are a combination of economic damage and lives lost. Parameters are chosen to be plausible for the British case. The presentation will demonstrate how control theorists can contribute to the pandemic problem currently facing the whole world. It will introduce the standard model used in epidemiology, and discuss some variations. It will then show how the problem can be posed as an optimal control problem, and how it can be solved. A trade-off between economic damage and the number of deaths is obtained, and a surprising discontinuity is demonstrated. We consider the effects of introducing “test and trace” capability, and the effect of delay in introducing social restrictions. We also discuss how this approach should be modified for real-time policy making, and note some other work on Covid-19 by other Control researchers.

This talk is based on the recently-published paper “A cost-benefit analysis of the Covid-19 disease” by Robert Rowthorn and Jan Maciejowski, published in the Oxford Review of Economic Policy, special issue on the Covid-19 pandemic.

Please join this talk using the following Zoom link:

How do our brains form maps of the world? – Alexandra Constantinescu (University College London) @ Zoom
Oct 22 @ 2:00 pm – 3:00 pm

Abstract not available

Topological Proximity Effects – Karyn Le Hur, CNRS and Ecole Polytechnique @ Details of video conferencing will be distributed nearer the time.
Oct 22 @ 2:00 pm – 3:00 pm

We discuss the induction of new topological band structures from proximity effects and entanglement. We present how to engineer a p+ip topological superconductor in quantum wires with spin-orbit coupling from the vicinity of a reservoir with s-wave Cooper pairing. Topological properties of energy bands associated with a spin-1/2 representation can be characterized from the reciprocal space and the Bloch sphere. The topological Chern number of one spin-1/2 model is known to be an integer. We highlight how to realize fractional topological phases directly from the reciprocal space, through the proximity effect and entanglement between two Bloch spheres. We show the applicability of these rational-valued Chern numbers in quantum circuits and a correspondence with honeycomb bilayer systems developing a nodal semimetallic ring at one Dirac point only.

Why is Covid such a wicked problem Bernard Silverman – Chris Dye – Andrew Dobson –
Oct 22 @ 2:00 pm – 4:00 pm
Linking the global and the local: the double burden of child malnutrition in Jamaica, c. 1960–2020 – Henrice Altink (University of York) @ Zoom
Oct 22 @ 3:30 pm – 5:00 pm

Following independence in 1962, successive governments in Jamaica tried to reduce the high rate of child malnutrition. Malnutrition was the result of a lack of protein and calories, also called PCM – Protein Calorie Malnutrition – and was a leading cause of death. Since the 1990s, however, the island has witnessed a nutrition transition with child malnutrition declining and child obesity increasing. Based on, amongst others, medical journals, newspaper reports, ministry papers, and reports of international agencies, this paper first of all explores how child malnutrition was measured and analysed; the various proposals put forward and implemented to reduce it; and the success rate of these policies. It will show that over time child malnutrition and the solutions proposed became increasingly localised; that is, greater attention was paid to the socio-economic and cultural context of pre-school children and their families and there was less reliance on outside agencies to reduce PCM. The paper will then move on to trace the rise in child obesity levels and show that contrary to the UK, US and many other western countries, child obesity in Jamaica is largely associated with higher income groups. Although child obesity has rapidly increased – in 2017 some 10.3% of children were obese – very few attempts have so far been made to localise the problem. The paper will explain why only recently campaigns – both government and NGO funded – have been started to address child obesity.

Rethinking the QCD axion – Luca di Luzio (DESY) @ Virtual Seminar
Oct 22 @ 4:00 pm – 5:00 pm

Four decades after its prediction the axion remains the most compelling solution to the strong CP problem and a well motivated dark matter candidate, inspiring several ultrasensitive experiments based on axion-photon mixing. After reviewing the axion solution of the strong CP problem and the experimental landscape of axion searches, I will focus on some recent developments in axion model building suggesting that the QCD axion parameter space is much larger than what traditionally thought. The implications for astrophysical limits and current experiments will be discussed as well.

Self-healing cut-off wall materials – Benyi Cao, PhD Research Student at the Department of Engineering, University of Cambridge @ Delivered online via Zoom
Oct 22 @ 4:00 pm – 4:45 pm

The contamination of soil and groundwater is a widespread issue that can adversely affect human health and ecosystems. The containment approach using cut-off walls is one of the most commonly used land remediation technologies. Despite the widespread application of cement-bentonite slurry trench cut-off walls and soil mix technology constructed cut-off walls, the materials deteriorate under mechanical, chemical, and environmental stresses. The damage of cut-off walls can lead to the problems related to the undermined mechanical and transport properties, impacting their serviceability and reliability and in some cases leading to the undetected physical breach of the wall. Developing crack-resistant and self-healing cement-bentonite and soil mix materials could provide more resilient, sustainable, and reliable cut-off walls with significantly enhanced durability, reduced maintenance costs, enhanced safety, and protection against sudden or undetected failure. This PhD research investigates the development of crack-resistant and self-healing cementitious cut-off wall materials incorporating polymers and minerals. The overall performance of four additives, including superabsorbent polymers (SAPs), oil sorbent polymers, reactive magnesia (MgO) and microencapsulated sodium silicate (Na₂SiO₃) in two cut-off wall materials (cement-bentonite slurry and soil mix cut-off wall materials) is the interest of this study.

Estimation – SEJ – Phillip Dawid and Will Probert –
Oct 23 @ 2:00 pm – 4:00 pm
Is the Halley temperature record homogeneous? – Prof. John King, British Antarctic Survey @
Oct 23 @ 2:00 pm – 3:00 pm

Commencing in 1956, observations made at Halley Research Station provide one of the longest continuous series of near-surface temperature observations from the Antarctic continent. The record does not, however, come from a single location but is a composite of observations from a sequence of seven stations, all situated on the Brunt Ice Shelf, that range from around 10 km to 50 km distance from the ice shelf coast. Until recently it was generally assumed that temperature data from all of these stations could be combined into a single composite record with no adjustment. In this talk I will report on recent research that suggests that this assumption of homogeneity is incorrect. Application of an objective statistical change-point detection algorithm indicates that there is at least one sudden downward jump in temperature associated with relocation of the station, which is large enough to introduce a spurious cooling trend into the composite record. Analysis of observations from a network of automatic weather stations and data from a run of a high-resolution regional atmospheric model confirm the existence of mean temperature gradients across the ice shelf which are large enough to explain the jumps seen in the record. These temperature gradients are largely driven by the advection of maritime air masses across the cold surface of the ice shelf. I will discuss the implications of these findings for local and regional climate studies that make use of the Halley record and will consider whether it might be possible to create a homogenised record.

New 1D models for localisation in slender structures – Claire Lestringant, Structural Engineering, Cambridge @ zoom 853 4440 7455
Oct 23 @ 2:00 pm – 3:00 pm

Slender structures are subject to various localised instabilities: necking of bars under traction, bulging of cylindrical party balloons, beading in cylinders made up of soft gels, or folding of tape-springs. In all these examples, distinct states of deformation may coexist and classical one-dimensional (1D) models predict singular solutions. In particular, classical 1D models fail to describe interfaces or finite size effects. The most common remedy is to use full structural models based on 3D finite elasticity or nonlinear shell/membrane equations. However, this is computationally costly and often impracticable: simpler 1D regularised models depending on the strain and the strain gradient are therefore attractive.
There is a recent effort to rigorously establish 1D higher-order models for the analysis of localisation in slender structures. I will introduce a systematic method to derive such models by a formal expansion, starting from a variety of full structural models for slender elastic structures. The expansion is performed near a finitely pre-strained state and therefore retains all sources of nonlinearity, coming from the geometry and the constitutive law. I will illustrate the method in the case of bulging and beading, and demonstrate its accuracy by comparing solutions of the 1D gradient model with solutions of the original structural model.

Sustainability as a Structural Engineer – Past, Present, Future – Natasha Watson, Buro Happold @ Zoom (email for link)
Oct 23 @ 3:00 pm – 4:00 pm

will discuss what sustainability
means to the practicing structural engineer
and our increasing role as leading change in
the built environment In the past,
sustainable design was about energy
efficiency and building out of timber Now we
realise it is much more than that, as can be
seen in the Construction Declares manifesto
In order to tackle the Climate Emergency we
need to build less, refurbish wherever
possible, and be considered and thoughtful
about where we use our carbon Natasha
will talk through some of the initiatives and
pragmatic ways in which she is reducing the
impact her designs have on climate change,
for example challenging the client on what
‘ really means, and understanding
where ‘quick win’ carbon savings can be
made if programme and fee are tight

A system integration route towards resonant MEMS sensing – Malar Chellasivalingam, CUED @ Online
Oct 23 @ 4:00 pm – 4:30 pm

Abstract not available