skip to content

NanoManufacturing

Michael De Volder, Engineering Department - IfM
 

Fri 07 Feb 12:00: LLMs as supersloppers and other metaphors

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 12:48
LLMs as supersloppers and other metaphors

Abstract: The interdisciplinary pilot project `Exploring novel figurative language to conceptualise Large Language Models’ is funded by Cambridge Language Sciences. This talk mainly concerns `slop’, by which we mean text delivered to a reader which is of little or no value to them (or is even harmful) or is so verbose or convoluted that the value is hidden. Examples range from unnecessarily obscure legal agreements to academic papers in predatory journals. Intuitively, slop has negligible information density for the reader. Although the term `slop’ is often used specifically for generative AI output, we argue that slop predates the machine-generation of text and has to be understood in the context of wider societal processes. Modern ways of delivering text are driving an increase in slop. In this context, systems incorporating LLMs have the capacity to be `supersloppers’—- tools for the creation and delivery of more and more pointless text.

Depending on time, I will discuss some other examples of figurative language and other ways of exploring aspects of LLMs. This will be a very informal talk – I am especially keen to hear about metaphors that other people find helpful (or unhelpful).

Ann Copestake, Lucy Duggan, Aurelie Herbelot, Amira Moeding and Eva von Redecker (2024). `LLMs as supersloppers’ Cambridge Open Engage. doi:10.33774/coe-2024-dx12p

Add to your calendar or Include in your list

Wed 28 May 11:15: Title to be confirmed

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 12:46
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Defect-rich Carbon Induced Built-in Interfacial Electric Field Accelerating Ion-conduction towards Superior-stable Solid-state Batteries

http://feeds.rsc.org/rss/ee - Tue, 04/02/2025 - 12:36
Energy Environ. Sci., 2025, Accepted Manuscript
DOI: 10.1039/D4EE05966B, PaperLiyu Du, Yiming Zhang, Yiyang Xiao, Du Yuan, Meng Yao, Yun Zhang
The electrochemical performances of composite solid-state electrolytes (CSEs) cannot satisfy the application requirements of solid-state batteries (SSBs) due to the low concentration of movable cations with disordered and slow cation...
The content of this RSS Feed (c) The Royal Society of Chemistry

Thu 06 Feb 17:00: Formalisation of Combinatorial Optimisation in Isabelle/HOL: Network Flows

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 11:51
Formalisation of Combinatorial Optimisation in Isabelle/HOL: Network Flows

Combinatorial optimisation (CO) is a sub-area of discrete mathematics. Basic examples for CO problems are finding a shortest path or a minimum spanning tree in a graph. So-called network flows or variations of matching would be more advanced problems. There are also abstract concepts like matroids that offer an algebraic point of view and a uniform foundation for some of the more concrete problems.

Since the considered structures are finite, it is a natural aim to compute a solution efficiently. That implies an overlap with the theory of algorithms, especially running time analysis.

This talk is mainly about the Isabelle/HOL formalisation of a specific CO problem, namely, minimum cost flows, which are a subtype of network flows. Among others, this includes Orlin’s Algorithm, which is a most efficient method to compute a minimum cost flow in general networks. Also, the running time argument for this advanced algorithm and some reductions among flow problems were formalised.

- The Isabelle proof scripts can be found in this GitHub repo: https://github.com/mabdula/Isabelle-Graph-Library

- The formalisation is described in this paper: A Formal Analysis of Capacity Scaling Algorithms for Minimum Cost Flows by Mohammad Abdulaziz and Thomas Ammer, ITP 2024

=== Hybrid talk ===

Join Zoom Meeting https://cam-ac-uk.zoom.us/j/87143365195?pwd=SELTNkOcfVrIE1IppYCsbooOVqenzI.1

Meeting ID: 871 4336 5195

Passcode: 541180

Add to your calendar or Include in your list

Fri 14 Mar 10:30: TBC (Mini-course talk 2) Please note the unusual time and place.

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 11:41
TBC (Mini-course talk 2)

Abstract not available

Please note the unusual time and place.

Add to your calendar or Include in your list

Thu 13 Mar 11:15: TBC (Mini-course talk 1) Please note the unusual time and place.

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 11:40
TBC (Mini-course talk 1)

Abstract not available

Please note the unusual time and place.

Add to your calendar or Include in your list

Tue 04 Feb 14:00: Researchers‘ experiences with vulnerability disclosures

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 11:38
Researchers‘ experiences with vulnerability disclosures

Vulnerabilities are becoming more and more prevalent in scientific research. Researchers usually wish to publish their research and, before that, have the vulnerabilities acknowledged and fixed, contributing to a secure digital world. However, the vulnerability disclosure process is fraught with obstacles, and handling vulnerabilities is challenging as it involves several parties (vendors, companies, customers, and community). We want to shed light on the vulnerability disclosure process and develop guidelines and best practices, serving vulnerability researchers as well as the affected parties for better collaboration in disclosing and fixing vulnerabilities.

We collected more than 1900 research papers published at major scientific security conferences and analyzed how disclosures are reported, finding inconsistent reporting, as well as spotty acknowledgments and fixes by affected parties. We then conducted semi-structured interviews with 21 security researchers with a broad range of expertise who published their work at scientific security conferences and qualitatively analyzed the interviews.

We discovered that the main problem starts with even finding the proper contact to disclose. Bug bounty programs or general-purpose contact email addresses, often staffed by AI or untrained personnel, posed obstacles to timely and effective reporting of vulnerabilities.

Experiences with CERT (entities supposed to help notify affected parties and facilitate coordinated fixing of vulnerabilities) were inconsistent, some extremely positive, some disappointing. Our interviewees further talked about lawsuits and public accusations from the vendors, developers, colleagues, or even the research community. Successful disclosures often hinge on researcher experience and personal contacts, which poses personal and professional risks to newer researchers.

We’re working on making our collected best practices and common pitfalls more widely known both to researchers and industry, for more cooperative disclosure experiences.

Zoom link: https://cam-ac-uk.zoom.us/j/89699287551?pwd=shaVGdAyVagZX2AvrVI9mazeKk8ssI.1

Meeting ID: 896 9928 7551 Passcode: 471680

Bio: Yasemin Acar (she/her) is a professor of computer science at Paderborn University, Germany, and a research assistant professor at The George Washington University. She focuses on human factors in computer security. Her research centers humans, their comprehension, behaviors, wishes and needs. She aims to better understand how software can enhance users’ lives without putting their data at risk. Her recent focus has been on human factors in secure development, investigating how to help software developers implement secure software development practices. Her research has shown that working with developers on these issues can resolve problems before they ever affect end users. Her research has won distinguished paper awards at IEEE Security and Privacy and USENIX Security, as well as a NSA best cyber security paper competition. Her web page: https://yaseminacar.de.

Add to your calendar or Include in your list

Wed 12 Mar 16:00: TBC (Geometry Colloquium)

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 11:38
TBC (Geometry Colloquium)

Abstract not available

Add to your calendar or Include in your list

Tue 04 Feb 13:00: Text-and-audio methods

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 11:00
Text-and-audio methods

This talk supports the R255 Advanced Topics in Machine Learning module on Multimodal Learning and provides a bird’s eye view of the rapidly evolving text-audio landscape, with a focus on music as a primary example of audio data. I will first present types of tasks that exist in this space, then discuss data curation challenges and follow with an overview of some existing retrieval and generation methods, including a quick primer on diffusion models. Finally, I will describe current evaluation metrics and their limitations.

You can also join us on Zoom

Add to your calendar or Include in your list

Wed 14 May 11:15: Title to be confirmed

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 10:54
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Wed 19 Mar 11:15: Title to be confirmed

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 10:53
Title to be confirmed

Abstract not available

Add to your calendar or Include in your list

Wed 05 Mar 11:15: Emergent phenomena in nanosculpted devices of quantum materials

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 10:52
Emergent phenomena in nanosculpted devices of quantum materials

Electrons typically traverse a conductive medium in a diffusive manner, resulting in a linear relationship between the measured voltage and applied current – known as Ohm’s law. However, violations of Ohm’s law may be found when the inherent symmetries of the underlying system are broken. Examples include the sliding motion of density waves; ballistic or hydrodynamic electron transport; or the symmetry-breaking realised by lattice or magnetic order. Focused ion beam (FIB) fabrication methods enable precise nanoscale devices to be fashioned from high-quality single crystalline materials, ideal for exploring these nonlinear phenomena. Such nanoengineering offers vast potential for the investigation of both fundamental physics and the development novel quantum devices. In this talk, I will introduce three specific examples. Firstly, we will explore the current-induced sliding motion of a skyrmion lattice in Gd2PdSi3 and the resulting emergent electrodynamics, which originate from a time-dependent Berry phase. Secondly, I will highlight our latest breakthrough to develop FIB fabrication of three dimensional nanostructures, in the form of helical-shaped devices of the high-mobility Weyl magnet CoSn2S2. By breaking inversion symmetry on the length scale of the electron mean free path, we observe large nonreciprocal transport, resulting in a switchable diode effect. Finally, if time permits, I will discuss the possibility to fabricate highly symmetrical devices, which allows the probing of symmetry breaking along multiple directions of a material simultaneously – in this case exploited to study signatures of p-wave magnetism in Gd3Ru4Al12.

Add to your calendar or Include in your list

Thu 27 Mar 09:30: Revealing the Unseen: AI's Role in Novel Target Discovery for High Unmet Needs Areas Through Multi-Omics Integrationce for identifying novel therapeutic targets, biomarkers and drug repositioning opportunities

http://talks.cam.ac.uk/show/rss/5408 - Tue, 04/02/2025 - 10:50
Revealing the Unseen: AI's Role in Novel Target Discovery for High Unmet Needs Areas Through Multi-Omics Integrationce for identifying novel therapeutic targets, biomarkers and drug repositioning opportunities

The role of AI and data-driven methodologies is set to play a huge part in the drug discovery industry, allowing intricate details of a multitude of disease mechanisms and improving the way for the streamlined development of therapeutics. The focus of this seminar would give an overview of how AI and Multi-Omics are shaping betterment in the area of the improved identification and profiling of therapeutic targets.

I will describe the general AI principles and show how they apply to drug discovery in a manner that can be followed by a wide audience. This should give the participants with even the barest understanding of the technical background an understanding of the value that this tool adds in drug discovery. I shall take our research using AI and machine learning coupled with advanced mathematical models. Integrative analysis of multi-omics datasets enabled identifying new targets and pathways that may suggest new avenues for their use in understanding complex biological questions.

The presentation also focuses on how such academic insights work out in the domain of commerce and further reflected through my experiences of founding and running AI-driven Startups: Kure.ai Therapeutics and CardiaTec Biosciences. The second one will bring out examples that evidently show the relevance of our findings in practical scenarios of developing innovative solutions in the area of cardiovascular diseases.

The talk concludes with a Q&A session and gives a very succinct but also very full picture of the role of AI and data-driven methods in modern drug discovery. It is in this regard that the subject of the seminar will put forth how these emerging technologies are enablers for increased precision and efficiency in the development of therapeutics today or will be, thus allowing an informed discussion among participants hailing from academia to biotech to big pharmaceuticals.

Add to your calendar or Include in your list

Expanding the Temperature Range for Stable Aqueous Batteries: Strategies, Mechanisms and Perspectives

http://feeds.rsc.org/rss/ee - Tue, 04/02/2025 - 10:44
Energy Environ. Sci., 2025, Accepted Manuscript
DOI: 10.1039/D4EE05304D, Review ArticleXianwei Fu, Ruijuan Shi, Ye Liu, Xiaoxiao He, Qian Li, Yan Zhang, Yong Zhao, Shilong Jiao
Aqueous batteries (ABs) based on water-containing electrolytes are intrinsically safe and serve as promising candidates for the grid-scale energy storage and power supplies of wearable electronics. The severe temperature fluctuations...
The content of this RSS Feed (c) The Royal Society of Chemistry

Scalable Copper Current Collectors with Precisely Engineered Lithiophilic Alloy “Skins” for Durable Lithium-Metal Batteries

http://feeds.rsc.org/rss/ee - Tue, 04/02/2025 - 10:44
Energy Environ. Sci., 2025, Accepted Manuscript
DOI: 10.1039/D4EE05862C, Paper Open Access &nbsp This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence.Huiqun Wang, Yuxiang Mao, Peng Xu, Yu Ding, Huiping Yang, Jian-Feng Li, Yu Gu, Jiajia Han, Li Zhang, Bingwei Mao
Depositing a uniform lithium metal layer on a highly conductive current collector (CC) is essential for the development of next-generation Li metal batteries (LMBs). However, poor cycling stability, low Coulombic...
The content of this RSS Feed (c) The Royal Society of Chemistry

Sodium cluster-driven safety concerns of sodium-ion batteries

http://feeds.rsc.org/rss/ee - Tue, 04/02/2025 - 10:44

Energy Environ. Sci., 2025, Advance Article
DOI: 10.1039/D4EE05509H, PaperJiaping Niu, Junyuan Dong, Xiaohu Zhang, Lang Huang, Guoli Lu, Xiaolei Han, Jinzhi Wang, Tianyu Gong, Zheng Chen, Jingwen Zhao, Guanglei Cui
This study reveals that quasi-metallic sodium clusters in hard carbon possess an electron state density even higher than that of metallic sodium, making them more susceptible to triggering thermal runaway compared to lithium in graphite.
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

Repairing the interfacial defect via preferable adsorption of ytterbium enables high-utilization and dendrite-free Zn metal anode

http://feeds.rsc.org/rss/ee - Tue, 04/02/2025 - 10:44
Energy Environ. Sci., 2025, Accepted Manuscript
DOI: 10.1039/D4EE05382F, PaperLong Jiang, Zhenyue Xing, Yanfen Liu, Xiaodong Shi, Le Li, Yangyang Liu, Bingan Lu, Jiang Zhou
Dendrite growth and spontaneous corrosion of zinc (Zn) metal anodes pose significant challenges for their application in grid-scale energy storage, primarily due to the instability of the bulk phase characterized...
The content of this RSS Feed (c) The Royal Society of Chemistry

A Highly Efficient Electrolysis System Enabled by Direct Impedance Matching Between Charge Migration Triboelectric Nanogenerator and Series Connected Electrolysers

http://feeds.rsc.org/rss/ee - Tue, 04/02/2025 - 10:44
Energy Environ. Sci., 2025, Accepted Manuscript
DOI: 10.1039/D4EE05522E, PaperYu Deng, Qian Qin, Wencong He, Hengyu Guo, Jie Chen
As an electromechanical conversion technology, triboelectric nanogenerators (TENGs) are widely used in water electrolysis for hydrogen production. Nevertheless, the impedance mismatch between TENGs and conventional electrolysers significantly reduces energy utilization...
The content of this RSS Feed (c) The Royal Society of Chemistry

Latest news

We are hiring!

4 January 2021

We are seeking to hire a research assistant to work on carbon nanotube based microdevices. More information is available here: www.jobs.cam.ac.uk/job/28202/

We are Hiring!

4 January 2021

We are seeking to hire a postdoc researcher to work on the structuring of Li-ion battery electrodes. More information is available here: www.jobs.cam.ac.uk/job/28197/