@filippocmc

Psychiatrist | IPPRF Fellow @imperialcollege - Digital Mental Health

 London
Joined March 2015
  • Tweets72
  • Following110
  • Followers80
5Media
and congratulations to Dr. @filippocmc who successfully passed his viva, examined by Paolo Ossola and @KiaNazarpour Filippo's thesis is a wonderful bridge between #ML #AI and clinical practice in psychiatry for mood disorders with wearable!
Filippo Corponi@filippocmc24 Jul 2024
1/2 If you are working in #wearables, #AI, and #healthcare do not let small datasets stop you. mhealth.jmir.org/2024/1/e550… 👈 We share the largest publicly available data collection for #empatica #e4 and release the codebase for pre-processing and self-supervised pre-training.

Wearable Data From Subjects Playing Super Mario, Taking University Exams, or Performing Physical...

Background: Personal sensing, leveraging data passively and near-continuously collected with wearables from patients in their ecological environment, is a promising paradigm to monitor mood disorders...

mhealth.jmir.org
1
6
40
2,685
#NeuroSymbolic #AI 🤖 enforces constraints, but models can achieve high accuracy using wrong concepts. Can we spot when a model relies on flawed concepts and ensure #trustworthiness? Yes, with BEARS! 🐻 📢 Introducing our latest paper, accepted as a spotlight at UAI2024 !🎉📄
1
8
2
38
6,628
we evaluate #continual learning models on "baby" benchmarks, where it is easy to show no catastrophic forgetting! we propose a new simple benchmark that glues simple but still challenging tasks in a curriculum: from MNIST to Imagenet and back! 📜link.springer.com/article/10…
1
15
1
53
7,499
Will be at AISTATS this week, would love to chat if you have integrals to approximate or are generally into compstat / opt. transport / estimation in causal inference. Also we have two papers around importance sampling and variational inference. 1st: adaptive IS for heavy tails
2
17
1
48
9,900
Filippo Corponi retweeted
Andreas Grivas@andreasgrv16 Apr 2024
The softmax bottleneck is an interesting problem; it has many side effects which we do not yet fully understand! If you want to build an intuition for the problem, here is an interactive visualisation I made grv.unargmaxable.ai/static/f… (best viewed on desktop).
Nathan Godey @COLM 🌉@nthngdy15 Apr 2024
Replying to @nthngdy
(6/10) This problem is in fact very much related with the softmax bottleneck issue (arxiv.org/abs/1711.03953) Basically, we try to map "low" dimensional contextual representations to potentially high-dimensional contextual probability manifolds, using a simple linear layer:
2
23
1
138
15,779
Filippo Corponi retweeted
Michael Stam@mjstam16 Apr 2024
New preprint out!🚨We performed a large scale analysis of physicochemical features extracted from over half a million AlphaFold structural models, and using data-driven methods we showed a link between these features and in vivo behaviour of proteins. Find out more below⬇️1/9
bioRxiv@biorxivpreprint14 Apr 2024
Large scale analysis of predicted protein structures links model features to in vivo behaviour biorxiv.org/cgi/content/shor… #bioRxiv
1
10
25
5,756
Filippo Corponi retweeted
Pedromics@pedromics9 Apr 2024
Prepare the pentobarbital.
2
98
8
621
48,069
Classical mixture models are limited to positive weights and this requires learning very large mixtures! Can we learn (deep) mixtures with negative weights? Answer in our #ICLR2024 spotlight by @loreloc_ Aleks, Martin, Stefan, Nicolas @arnosolin 📜openreview.net/forum?id=xIHi…
Jakub Tomczak@jmtomczak9 Apr 2024
After the score-based models, I decided to take a step back and cover the basics in a new blog post: Probabilistic modeling and Mixture Models. Additionally, I make a brief intro to ✨Probabilistic Circuits✨ Check: 📄Post: jmtomczak.github.io/blog/19/… 🖥️ Code: github.com/jmtomczak/intro_d…
4
44
4
216
53,831
Filippo Corponi retweeted
Andreas Grivas@andreasgrv16 Feb 2024
How expressive is your deep multi-label classifier? Can it represent all outputs of interest? 🤔 SPOILER🚨: Your model can have *test set* outputs that are impossible to predict! 🚫 Check out our paper! arxiv.org/abs/2310.10443 🧵(1/7)
2
21
3
71
20,588
🎉 Exciting News! Our paper "TIMED-Design: Flexible & Accessible Protein Sequence Design with Convolutional Neural Networks" has been accepted for publication in Protein Engineering, Design and Selection! 👉 Try the UI: pragmaticproteindesign.bio.e… 📖 Paper: academic.oup.com/peds/advanc…
3
20
78
6,983
when they ask me why I won't be going back to #Italy anytime soon a Vatican priest with a PhD in moral theology and some interests in ethical aspects of AI is now leading the Italian task force of #AI (previously led by a 90 years old guy with no experience in computer science)
Ultimora.net - POLITICS
@ultimora_pol5 Jan 2024
Il teologo francescano Paolo #Benanti è il nuovo Presidente della Commissione sull'Intelligenza Artificiale per l'informazione, dopo le dimissioni di Giuliano #Amato @ultimora_pol
6
11
3
70
16,598
It still amazes me that data in #precision #psychiatry is so scarce, even for #wearables! This is why we put together 11 (small) datasets for #EmpaticaE4 creating the largest publicly available dataset (260+ days, 250+ subjects). Which you can use for self-supervised learning!
Filippo Corponi@filippocmc8 Nov 2023
Replying to @filippocmc
🧵5/5 We are soon to release the E4SelfLearning collection along with our pre-processing pipeline doing on-/off-body detection, sleep-wake detection, segmentation, and (optionally) feature extraction.
5
9
1,844
Filippo Corponi retweeted
Lorenzo Loconte@loreloc_2 Dec 2023
More reliable and efficient generative KGE models without compromises. GeKCs achieve the highest score on the OGB biokg benchmark when compared to single-model solutions in the last ~2 years, while requiring just ~1 hour of training time on a single GPU.
Lorenzo Loconte@loreloc_29 Nov 2023
We convert popular knowledge graph embedding (KGE) models such as ComplEx into generative models of triples that 1️⃣ better scale training to very large knowledge graphs 2️⃣ make reliable predictions with logical constraints 3️⃣ support sampling in our #NeurIPS2023 oral!
4
16
2,933
do you want to train an #embedding model on #wikidata on a single GPU in 2 hours (instead of 2 days)? use very large batch sizes and #generative #KG #circuits GeKCs 👇
Lorenzo Loconte@loreloc_29 Nov 2023
Replying to @loreloc_
While training KGE models is expensive (both in time and memory as shown in the GIF) when dealing with large knowledge graphs with millions of entities, GeKCs come with significant savings as they allow for a much more efficient evaluation of loss functions on GPUs.
4
16
2,291
Filippo Corponi retweeted
Pasquale Minervini@PMinervini4 Nov 2022
Interested in a PhD on NLP/ML-related topics with me at @EdinburghUni? 🧑‍🎓🚀 Apply to either the UKRI CDT in NLP (@Edin_CDT_NLP) or to the ILCC PhD Program - the deadline is around the corner! More at neuralnoise.com/2022/phd-pro… ⏰ International - Nov 25th, 2022 ⏰ UK - Jan 27th, 2023
UKRI CDT in Natural Language Processing@Edin_CDT_NLP27 Oct 2022
📢Applications welcome & encouraged from individuals relating to groups under-represented in technology for the FINAL cohort of UKRI CDT in NLP, @EdinburghUni! 🥹 Programme/Application Details: edin.ac/cdt-in-nlp Deadlines: ⏰Non UK – 25.11.22 ⏰ UK – 27.01.23 @InfAtEd
1
21
1
59
Filippo Corponi retweeted
Bryan M. Li@bryanlimy3 Nov 2023
Our spatial-temporal ViT approach to predict thousands of 🐭 V1 neural response to movie clips came🥉in the Sensorium challenge! Looking forward to presenting our solution at #NeurIPS2023! @wolfdewulf @NinelK1 @MatthiasHennig6 @RochefortLab Arno Onken @BioMedAI_CDT
Sensorium@SensoriumComp1 Nov 2023
The Sensorium 2023 competition is over. Here are the winning teams: 🥇IRomul 🥈YuZhu 🥉dunedin All winners used custom architectures. To learn more about their approaches, join our in-person workshop at NeurIPS on Friday, Dec 15, 2023. 🗓️ 1/4
1
15
31
5,326
Filippo Corponi retweeted
Bryan M. Li@bryanlimy18 Aug 2023
Happy to share that our (@isabelmaria_c, @RochefortLab, Arno Onken) paper "V1T large-scale mouse V1 response prediction using a Vision Transformer" is now published on @TmlrOrg. Paper: openreview.net/forum?id=qHZs… GitHub: github.com/bryanlimy/V1T 🧵
1
16
57
8,141
I have fully-funded PhD positions (3.5 yrs) for troublemakers in #ML #AI who want to design the next gen of #probabilistic #models and #programs that are provably #reliable and #efficient Join @InfAtEd @EdinburghUni Email me! ✉️ Please share! 🔁 Apply 👉nolovedeeplearning.com/buyse…
4
69
4
184
55,782
Looking forward to be teaching how #probabilistic models can be both #deep and #reliable! See you in Ljubiana 🇸🇮 this July!
Vida Groznik@TechVida9 May 2023
Learn Probabilistic Circuits from experts Robert Peharz and Antonio Vergari and improve your decision-making skills. Register now at essai.si. #essai23 #aisummerschool #machinelearning #probabilisticmodeling
6
35
2,988
5 days to #deepdive and learn about advanced #probabilistic #reasoning, #deep #generative models, integrating #logic #symbols #constraints and #deep #learning Please share 🙏
This Post is from an account that no longer exists. Learn more
6
11
2,473
Load more