Francesco Tudisco (@FFTudisco) header
Francesco Tudisco (@FFTudisco) profile picture

Machine Learning Professor @ Uni of Edinburgh | AI4Science | Numerics | HPC | Optimization | Co-founder @ https://teapawt.net/t.co/PgD3DKIOfj

Edinburgh, Scotland
Joined September 2016
Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)Photo by Francesco Tudisco (@FFTudisco)
The scientific community is transitioning to #bluesky and I am moving there too. See you on 🦋!
1
1
61
Francesco Tudisco retweeted
My understanding is that @OpenAI heavily use @WecoAI's codebase (AIDE, github.com/WecoAI/aideml) without giving proper credit -- they rebrand the benchmark as MLE-bench, use @WecoAI's evaluation metrics, and their contribution is ~limited to adding o1-preview on top of it 🫤
Excited to see OpenAI's recent project, MLE-bench, is based on our open-source effort, AIDE. In their independent evaluation, AIDE surpasses other MLE agents by a large margin!
4
32
3,338
Francesco Tudisco retweeted
Over 92% of authors of papers on Covid-19 knew nothing about infectious diseases! Problem or opportunity? Both! Read our comment, just out in @NatureHumBehav . @IULuddy @maiamajumder @FFTudisco @istudymemes @satyaki30 nature.com/articles/s41562-0…
6
1
35
2,744
Network science opportunity with fantastic supervisor @santo_fortunato 👇👇
Postdoctoral position on the science of science available in my group at Indiana University. Join us to explore knowledge landscapes, predict future discoveries, and more. Here is the ad: indiana.peopleadmin.com/post… @ICSSIConference @IULuddy
2
207
Francesco Tudisco retweeted
Working on this project was really fun and instructive! Also, Adit (@aadityaura) is insanely talented and currently looking for PhD positions -- catch him if you can! 🚀 (I'm available to help/co-sup via, e.g., @ELLISforEurope!)
Exciting news! 📢 In collaboration with Hugging Face 🤗, we are launching the Medical-LLM Leaderboard. This leaderboard aims to provide a standardized platform for evaluating large language models in the medical domain. It's encouraging to see tech companies like Google, Microsoft, and OpenAI, have already adopted our benchmarks to evaluate their models. With the support of Hugging Face, we are now making this leaderboard accessible to the wider community. 🔥 huggingface.co/blog/leaderbo…
1
5
1
27
6,653
Fantastic quote and fantastic analogy
A story by Luciano De Crescenzo (an Italian writer) lucianodecrescenzo.it/sono-l… At IBM in Naples there was a very slow elevator and people were complaining about waiting at the floors. However, substituting it was very expensive. Instead, they decided to put a mirror on each floor. So, people would kill time looking themselves at the mirrors and stopped complaining. The rebuttals are our mirrors, the elevator is the review system
2
255
Francesco Tudisco retweeted
Introducing AlphaGeometry: an AI system that solves Olympiad geometry problems at a level approaching a human gold-medalist. 📐 It was trained solely on synthetic data and marks a breakthrough for AI in mathematical reasoning. 🧵 dpmd.ai/alphageometry
113
924
378
3,813
1,197,698
Francesco Tudisco retweeted
Scientists can make their collaborators switch topics, especially if they are productive and highly cited! Check out our paper "Collaboration and topic switches in science", just out in Scientific Reports. @IULuddy @satyaki30 @FFTudisco rdcu.be/dvUcN
7
24
2,803
Francesco Tudisco retweeted
Adversarial Robustness of Artificial Intelligence - bit.ly/3QEoKON Call for papers to form a special issue of Eur. J. App. Math.
3
3
441
Francesco Tudisco retweeted
New blog post: Yet Another ICML Award Fiasco The story of the @icmlconf 2023 Outstanding Paper Award to the D-Adaptation paper with worse results that the ones from 9 years ago Please share it to start a needed conversation on mistakenly granted awards parameterfree.com/2023/08/30…
16
97
36
469
186,242
Francesco Tudisco retweeted
They proposed a workshop for #NeurIPS2012 on fairness in machine learning. It was rejected. Why? Not clear if anyone would be interested.
3
11
2
77
23,982
Francesco Tudisco retweeted
JOBS!!! Chancellor’s Fellowships (5-year tenure track lecturer/reader positions) now advertised in the School of Maths at the University of Edinburgh. Areas of interest: Data Science and AI; Climate and Environmental Sustainability; Health and Life. See: maths.ed.ac.uk/school-of-mat…
16
1
25
4,839
Francesco Tudisco retweeted
Mathematician at an AI conference.
1
26
3,295
Francesco Tudisco retweeted
The ILAS Early Career Speakers (a prize supported by @ElsevierConnect) in Madrid:  Teresa Conde, Andreas Bluhm, Massimiliano Fasi, Luyining Gan, Giovanni Barbarino! Many congratulations! #ILAS2023 #mathematics #linearalgebra @ElsevierNews
2
4
18
695
La risposta delle allieve e degli allievi della @scuolanormale al lutto nazionale
319
1,365
119
9,377
449,424
Excited to host Lorenzo Rosasco @lrntzrsc and Agnese Seminara @AgneseSeminara from @malga_center at @GSSI_LAQUILA for a fantastic lecture series on Learning and Navigating Dynamical Systems
1
6
1,287
Preprint alert for all hypergraph lovers (or haters): We introduce a method to learn whether we need more than pairwise interactions to represent the underlying structure of a dynamical system. Joint work with Michael Scholkemper @FFTudisco Michael Schaub arxiv.org/abs/2306.01813
8
1
28
1,938
Francesco Tudisco retweeted
Proud that our work on how collaborators push scientists to switch topics has been selected for a plenary talk at @IC2S2 in Copehagen. Fun collaboration with Sara Venturini, @satyaki30, Francesco Rinaldi and @FFTudisco. The paper is just out on the arXiv: arxiv.org/abs/2304.06826
4
21
2,008
Francesco Tudisco retweeted
Not long ago, breakthroughs in AI research often came from lone academics or small teams using desktop hardware. These days, not so much. Are you anxious about how to stay competitive in AI as an academic? @yannakakis and I wrote this piece for you: arxiv.org/abs/2304.06035
23
279
60
1,168
270,324