Thu 10 Jul 15:50: From Score to Sound: Music Generation in the AI Era
The seminar “From Score to Sound: Music Production in the AI Era” will explore the evolution of AI-driven methods for automatic music generation, from the early use of Recurrent Neural Networks to the latest Foundational Models. The talk will examine the shift from symbolic score modelling to audio-centric and multimodal approaches. It will also highlight the technical and conceptual advancements that have enabled the development of transformer and diffusion-based models, paving the way for today’s prompt-to-music systems.
- Speaker: Francesco Bardozzo (University of Salerno)
- Thursday 10 July 2025, 15:50-17:20
- Venue: Computer Laboratory, William Gates Building, Room FW26.
- Series: Foundation AI; organiser: Pietro Lio.
Tue 08 Jul 15:50: Artificial Intelligence in Agrifood and Environment
Artificial Intelligence (AI) is expected to have a transformative impact on the natural sciences by enhancing modeling capabilities and improving the prediction of natural phenomena across multiple spatial and temporal scales. This talk will highlight the urgency of coordinated scientific and regulatory initiatives to ensure the sustainable development of our planet. It will also provide an overview of recent advances in AI-driven approaches within the environmental domain, with a particular focus on solutions for coastal and marine ecosystem monitoring. Finally, the presentation will offer ideas on potential future developments from a modeling perspective, underscoring emerging directions and opportunities for interdisciplinary research.
- Speaker: Prof. Antonino Staiano
- Tuesday 08 July 2025, 15:50-17:20
- Venue: Computer Laboratory, William Gates Building, Room FW26.
- Series: Foundation AI; organiser: Pietro Lio.
Wed 09 Jul 09:00: AI Governance and Regulation: Comparing EU, US, and China’s Models and global initiatives
AI governance is increasingly shaped by a complex interplay of normative approaches. While high-level principles such as fairness, transparency, accountability, and safety are widely recognized across governance frameworks, their implementation varies significantly. The growing geopolitical significance of AI has driven governments to develop distinct strategies and policies, giving rise to 3 main models of AI governance. The Neoliberal Model, championed by the United States, prioritizes market-driven innovation, industry self-regulation, and minimal government intervention. Digital Sovereignty, exemplified by China, reflects a state-controlled and security-driven approach that emphasizes data localization and algorithmic transparency tailored to government priorities, particularly in information control and social stability. The European Union’s Digital Constitutionalism model embeds fundamental rights and democratic oversight into AI regulation, aiming for human-centric, trustworthy, and accountable AI governance. However, the boundaries between these governance paradigms are increasingly blurring. Under the Biden administration, the U.S. briefly moved closer to the EU model before reverting to a neoliberal stance, leveraging Big Tech as proxies of power and security actors. The EU struggles to balance its ambition to lead in Trustworthy AI with competitiveness and security concerns. China, while maintaining strict state control, has introduced selective innovation incentives and consumer rights protections with distinct “Chinese characteristics.” Rather than fostering a cross-fertilization of these models, these shifting boundaries appear to reflect escalating geopolitical tensions, making international consensus on AI governance increasingly difficult to achieve.
- Speaker: Dr. Nicola Palladino, University of Salerno
- Wednesday 09 July 2025, 09:00-10:30
- Venue: Computer Laboratory, William Gates Building, Room FW26.
- Series: Foundation AI; organiser: Pietro Lio.
Wed 09 Jul 09:00: AI Governance and Regulation: Comparing EU, US, and China’s Models and global initiatives
AI governance is increasingly shaped by a complex interplay of normative approaches. While high-level principles such as fairness, transparency, accountability, and safety are widely recognized across governance frameworks, their implementation varies significantly. The growing geopolitical significance of AI has driven governments to develop distinct strategies and policies, giving rise to 3 main models of AI governance. The Neoliberal Model, championed by the United States, prioritizes market-driven innovation, industry self-regulation, and minimal government intervention. Digital Sovereignty, exemplified by China, reflects a state-controlled and security-driven approach that emphasizes data localization and algorithmic transparency tailored to government priorities, particularly in information control and social stability. The European Union’s Digital Constitutionalism model embeds fundamental rights and democratic oversight into AI regulation, aiming for human-centric, trustworthy, and accountable AI governance. However, the boundaries between these governance paradigms are increasingly blurring. Under the Biden administration, the U.S. briefly moved closer to the EU model before reverting to a neoliberal stance, leveraging Big Tech as proxies of power and security actors. The EU struggles to balance its ambition to lead in Trustworthy AI with competitiveness and security concerns. China, while maintaining strict state control, has introduced selective innovation incentives and consumer rights protections with distinct “Chinese characteristics.” Rather than fostering a cross-fertilization of these models, these shifting boundaries appear to reflect escalating geopolitical tensions, making international consensus on AI governance increasingly difficult to achieve.
- Speaker: Dr. Nicola Palladino, University of Salerno
- Wednesday 09 July 2025, 09:00-10:30
- Venue: Lecture Theatre 2, Computer Laboratory, William Gates Building.
- Series: Foundation AI; organiser: Pietro Lio.
Exploiting strained epitaxial germanium for scaling low-noise spin qubits at the micrometre scale
Nature Materials, Published online: 06 July 2025; doi:10.1038/s41563-025-02276-w
The authors measure electric and magnetic noise, an important source of decoherence for quantum devices, on hole spin qubit devices in quantum wells in Ge/SiGe heterostructures, revealing a reduced charge noise on devices fabricated on Ge wafers.Wed 29 Oct 14:00: title tbc
Abstract not available
- Speaker: Prof. Dr. Ryan Gilmour, University of Münster,
- Wednesday 29 October 2025, 14:00-15:00
- Venue: Dept. of Chemistry, Wolfson Lecture Theatre.
- Series: Synthetic Chemistry Research Interest Group; organiser: Matt Gaunt.
Mon 20 Oct 14:00: title tbc
Abstract not available
- Speaker: Prof. Syuzanna Harutyunyan, University of Groningen.
- Monday 20 October 2025, 14:00-15:00
- Venue: Dept. of Chemistry, Wolfson Lecture Theatre.
- Series: Synthetic Chemistry Research Interest Group; organiser: Ruth Webster.
Mon 17 Nov 14:00: tbc
Abstract not available
- Speaker: Prof. Harry Anderson, University of Oxford
- Monday 17 November 2025, 14:00-15:00
- Venue: Dept. of Chemistry, Wolfson Lecture Theatre.
- Series: Synthetic Chemistry Research Interest Group; organiser: Chris Hunter.
Fri 04 Jul 15:50: Title to be confirmed
Abstract not available
- Speaker: Speaker to be confirmed
- Friday 04 July 2025, 15:50-17:20
- Venue: Lecture Theatre 2, Computer Laboratory, William Gates Building.
- Series: Foundation AI; organiser: Pietro Lio.
Fri 04 Jul 15:50: Title to be confirmed
Abstract not available
- Speaker: Speaker to be confirmed
- Friday 04 July 2025, 15:50-17:20
- Venue: Lecture Theatre 2, Computer Laboratory, William Gates Building.
- Series: Foundation AI; organiser: Pietro Lio.
Milliwatt-scale 3D thermoelectric generators via additive screen printing
DOI: 10.1039/D5EE01151E, Paper Open Access   This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.Sairam Antharam, Muhammad Irfan Khan, Leonard Franke, Zirui Wang, Nan Luo, Jan Feßler, Wenjie Xie, Uli Lemmer, Md Mofasser Mallick
This study reports an additive screen-printing method for a high-performance fully printed planar 3D TEG, achieving a milliwatt-scale power output of 1.22 mW at ΔT of 43 K, underlining its potential for battery-free IoT applications.
To cite this article before page numbers are assigned, use the DOI form of citation above.
The content of this RSS Feed (c) The Royal Society of Chemistry
Thu 10 Jul 13:00: Seminar cancelled - Spatial mapping of breast cancer tumour microenvironment in Black British and White British women **Seminar cancelled**
Women of Afro-Caribbean descent confront more aggressive breast cancer subtypes at a younger age than their Caucasian counterparts. Yet, breast cancer research and treatment development have predominantly focused on Caucasian populations, neglecting potential biological drivers of these disparities. Our study addresses this gap by in-depth characterising the breast tumour microenvironment (TME) in an ethnically diverse cohort. We analysed treatment-naïve breast cancer samples from 45 Black British and 45 White British women, matched by age, tumour subtype, and stage by employing spatial transcriptomics (NanoString GeoMx) and hyper-plex protein profiling (Leica Microsytems Cell DIVE ). We captured whole-transcriptome data from cancer (PanCK+), immune (CD45+), and stromal (aSMA+) compartments from both tumour centre and tumour edge. The most striking differences emerged within the immune and stromal compartments, not in the cancer cells, underscoring metabolic, adhesion, and extracellular matrix rewiring in Black British tumours. Complementary spatial protein profiling further revealed changes in tissue architecture with distinct recurrent patterns of cellular organisation and cell-cell interactions, involving endothelial and B-cells. Our findings suggest that the TME plays a pivotal role in driving ethnic disparities in breast cancer, highlighting the urgent need for ethnically tailored therapies and more inclusive clinical trials to advance precision cancer care. This breakthrough offers new avenues for improving overall outcomes in breast cancer.
**Seminar cancelled**
- Speaker: Professor Kairbaan Hodivala-Dilke, Centre for Tumour Microenvironment, 2 Barts Cancer Institute, a Cancer Research UK of Excellence, Queen Mary University of London, Charterhouse Square, London, EC1M 6BQ, UK
- Thursday 10 July 2025, 13:00-14:00
- Venue: CRUK CI Lecture Theatre.
- Series: Cancer Research UK Cambridge Institute (CRUK CI) Seminars in Cancer; organiser: Kate Davenport.
Thu 17 Jul 16:00: Elucidating T Cell Signalling Dynamics Using Reconstitution and Optogenetics
This Cambridge Immunology Network Seminar will take place on Thursday 17 July 2025, starting at 4:00pm, in the Ground Floor Lecture Theatre, Jeffrey Cheah Biomedical Centre (JCBC)
Speaker: Dr John James, Associate Professor, Immunology, Warwick Medical School, University of Warwick
Title: Elucidating T Cell Signalling Dynamics Using Reconstitution and Optogenetics
Abstract: T cells are an essential part of our immune system; they detect infected cells and either directly kill or orchestrate their removal to keep us healthy despite constant exposure to potential pathogens. Great progress has been made in identifying the parts of the signalling networks that T cells use to execute these decision-making processes, and we now have near-complete lists of these pathways. However, to fully describe T cell function we must also understand how signals traverse these network connections, but this knowledge remains far more limited in T cells.
To address this limitation, we use cellular reconstitution and light-mediated control over these signalling pathways to directly and quantitively investigate T cell signalling in the cellular context. In the talk, I will show how we have used these discovery-based tools to better understand the mechanisms of action for new therapeutics (bispecifics/CAR-T), as well as preliminary data on quantifying inhibitory receptor function. I will also present our reconstitution work on how the pre-T cell receptor can drive commitment to the αβ-T cell lineage in the absence of ligand.
Host: Mathilde Colombe and Tim Halim, CRUK Cambridge
Refreshments will be available following the seminar.
- Speaker: Dr John James, Immunology, Warwick Medical School. Warwick Medical School
- Thursday 17 July 2025, 16:00-17:00
- Venue: Lecture Theatre, Jeffrey Cheah Biomedical Centre, Cambridge Biomedical Campus.
- Series: Cambridge Immunology Network Seminar Series; organiser: Ruth Paton.
One ultrasonic measurement for non-invasive and whole-life-cycle thermal diagnosis of lithium-ion batteries
DOI: 10.1039/D5EE01892G, PaperLingshi Zhang, Zhongbao Wei, Chunxia Liu, Hongwen He, Kailong Liu, Guangmin Zhou, Yunhui Huang, Zhichuan J. Xu
Thermal characterization and diagnosis are critical for the whole-life-cycle safety of lithium-ion batteries (LIBs). However, conventional techniques are time-delayed and discontinuous due to the sealed structure and intricate mechanisms of...
The content of this RSS Feed (c) The Royal Society of Chemistry
Fri 19 Sep 08:45: Saving Lives Ethically: A Practical and Organisational Approach to Veterinary CPR.
Abstract not available
- Speaker: Malina Filipas, Department of Veterinary Medicine
- Friday 19 September 2025, 08:45-10:00
- Venue: LT2.
- Series: Friday Morning Seminars, Dept of Veterinary Medicine; organiser: Fiona Roby.
Interfacial Energetics Reversal Strategy for Efficient Perovskite Solar Cells
A general strategy- “interfacial energetics reversal” to reconstruct perovskite energetics that matches well with the upper hole transport layer has been successfully developed, enabling efficient n–i–p perovskite solar cells with nonradiative recombination induced qVoc loss of only 57 meV from the radiative limit.
Abstract
Reducing heterointerface nonradiative recombination is a key challenge for realizing highly efficient perovskite solar cells (PSCs). Motivated by this, a facile strategy is developed via interfacial energetics reversal to functionalize perovskite heterointerface. A surfactant molecule, trichloro[3-(pentafluorophenyl)propyl]silane (TPFS) reverses perovskite surface energetics from intrinsic n-type to p-type, evidently demonstrated by ultraviolet and inverse photoelectron spectroscopies. The reconstructed perovskite surface energetics match well with the upper deposited hole transport layer, realizing an exquisite energy level alignment for accelerating hole extraction across the heterointerface. Meanwhile, TPFS further diminishes surface defect density. As a result, this cooperative strategy leads to greatly minimized nonradiative recombination. PSCs achieve an impressive power conversion efficiency of 25.9% with excellent reproducibility, and a nonradiative recombination-induced qV oc loss of only 57 meV, which is the smallest reported to date in n-i-p structured PSCs.
Key Role of Bridge Adsorbed Hydrogen Intermediate on Pt–Ru Pair for Efficient Acidic Hydrogen Production
The designed Pt atomic chain modified fcc-type Ru nanocrystal with co-crystalline structure possesses an efficient bridge adsorbed hydrogen configuration (*Hbridge) at the Pt–Ru(fcc) interface. The *Hbridge dominates HER and exhibits superior intrinsic activity, ≈10.6 times higher than that of Pt. The designed Pt–Ru(fcc) with *Hbridge demonstrates excellent catalytic activity and stability for both laboratory and industrial levels.
Abstract
Atop and multiple adsorbed hydrogen are considered as key intermediates on Pt-group metal for acidic hydrogen evolution reaction (HER), yet the role of bridge hydrogen intermediate (*Hbridge) is consistently overlooked experimentally. Herein, a Pt atomic chain modified fcc-Ru nanocrystal (Pt–Ru(fcc)) is developed with a co-crystalline structure, featuring *Hbridge intermediate bonded on the Pt–Ru pair site. Electrons leap from the pair site to *Hbridge facilitate hydrogen desorption, thus accelerating the Tafel kinetics and ensuring outstanding electrocatalytic performance, with a low overpotential (4.0 mV at 10 mA cm−2) and high turnover frequency (56.4 H2 s−1 at 50 mV). Notably, the proton exchange membrane water electrolyzer PEMWE with ultra-low loading of 10 ugPt cm−2 shows excellent activity (1.61 V at 1.0 A cm−2) and low average degradation rate (4.0 µV h−1 over 1000 h), significantly outperforming the benchmark Pt/C. Furthermore, the PEMWE-based 80 µm Gore membrane under identical operating conditions requires only 1.54 and 1.58 V to achieve 1.0 and 1.5 A cm−2. This finding highlights the key role of *Hbridge at the Pt–Ru interface in obtaining high HER intrinsic activity and underscores the transformative potential in designing next-generation bimetallic catalysts for clean hydrogen energy.
Soft Extrudable Dendritic Particles with Nanostructured Tendrils for Local Adhesion and Drug Release to Bladder Cancers
Soft dendritic particles (SDPs) made of biodegradable polymers and small-molecule drugs are manufactured using a fluid flow templating method. In alginate gels, extruded SDPs sustainably adhere on tumor sites and selectively kill cancer cells. Intravesical instillation of SDPs in tumor-bearing mouse bladders triggers a CD45+ immune response with minimal toxicity, highlighting their potential for targeted cancer therapy.
Abstract
Bladder cancer is a leading cause of cancer-related mortality, yet current intravesical drug delivery methods often suffer from poor retention times in the bladder. Gecko feet-like nanomaterials offer the potential to overcome this challenge, however, conventional methods to fabricate high surface area nanomaterials for drug delivery involve complex and expensive manufacturing processes. In this work, a simple fluid flow templating method is reported for manufacturing soft dendritic particles (SDPs) composed of poly(lactic-co-glycolic acid) (PLGA) with a chitosan coating for enhanced adhesion to epithelial tissues via van der Waals interactions. The biodegradable SDPs encapsulate chemotherapeutic agents and are administered using an alginate hydrogel, enabling precise deposition by extrusion for sustained drug release. The results demonstrate that SDPs adhere to mouse and human cancer cells for several days. The SDPs effectively encapsulate and release several clinically utilized chemotherapeutic drugs such as gemcitabine, docetaxel, and methotrexate, exhibiting superior cancer cell killing in vitro. In murine models, gemcitabine-loaded SDPs instilled into tumor-bearing bladders elicited stronger CD45+ immune cell responses than control groups while maintaining minimal toxicity. This work presents a simple, biomimetic drug delivery platform with prolonged retention and controlled drug release, offering a versatile approach for enhancing therapeutic delivery in epithelial cancer models.
Valence and Site Engineering Enable Efficient Broadband Near‐Infrared Emission at 960 nm in Cr3+‐Activated Forsterite
This study reports a near-infrared (NIR) luminescent material (Mg2SiO4:1.5%Cr,5%Li) with efficient (EQE = 48%) ultra-broadband (FWHM ≈ 2413 cm−1 (226 nm)) NIR emission at 960 nm upon blue-light excitation. This innovation addresses the low efficiency of long-wavelength (λmax > 900 nm) NIR-emitting materials due to the radiationless de-activation and the low absorption efficiency of Cr3+ activators.
Abstract
Near-infrared (NIR) light sources hold great potential for applications in night vision illumination, bio-imaging, and non-destructive testing. However, radiationless de-activation and low absorption restrict the development of high-efficiency blue light excitable NIR phosphors, especially for emissions beyond 900 nm. Herein we report a high-performance Cr3+-activated forsterite (Mg2SiO4:1.5%Cr3+, 5%Li+) phosphor exhibiting broadband NIR emission peaking at 960 nm with a record external quantum efficiency (EQE) up to 48%. The introduction of Li+ as a charge compensator and symmetry distorter not only suppresses Cr4+ formation but also enhances the cross section of Cr3+ d-d forbidden transitions in Mg2SiO4. More importantly, Li+ promotes excited-state energy transfer between Cr3+ emitters, yielding exceptional thermal stability and external quantum efficiency. A fabricated NIR phosphor-converted light-emitting diode (LED) achieved a NIR radiated power of 356 mW (at a driving current of 700 mA) and an electro-optical conversion efficiency up to 12.9% (at 100 mA). This work unlocks new possibilities for smart spectroscopy applications, from non-destructive testing to human angiography and biometric recognition.
All-day freshwater and power generation via integrated photothermal-enhanced thermoelectrics and evaporation cooling
DOI: 10.1039/D5EE02663F, PaperWenhe Zhang, Chengbing Wang, Lu Wang, Fan Wang, Puxin Tan, Jinchi Ma, Jingjing Jin, Zhongrong Geng, Hongyao Xie, Li-Dong Zhao
Solar-powered simultaneous electricity and freshwater production is a promising solution to address energy and water shortages. However, current technologies are limited by their reliance on sunlight and have yet to...
The content of this RSS Feed (c) The Royal Society of Chemistry