Lorenzo Bertolini

Unserious post:

Now that (some) #AI researchers and engineers have rockstars-level salaries and notoriety, will they also have stage (or, better, paper) names?

Made it alive and well to the last day of #ACL2025NLP? Hats off you to you (well…us…)!!!

Time to collect the last energy and pass by our poster this afternoon 😀

Not pressing _this_ work at #ACL2025NLP, but always happy to chat NLP/AI for dream analysis 😴🤖

Strengths and Limitations of Word-Based Task Explainability in Vision Language Models: a Case Study on Biological Sex Biases in the Medical Domain Lorenzo Bertolini, Valentin Comte, Victoria Ruiz-Serra, Lia Orfei, Mario Ceresa. Proceedings of the 6th Workshop on Gender Bias in Natural Language Processing (GeBNLP). 2025.

Paper: aclanthology.org/2025.gebnlp-...

I’ll be at #ACL2025NLP next week, with a paper for the #GeBNLP workshop!

In this work, we tested the ability of a task-level explainability method to trace biological-sex biases in medical image classification, using general and biomedical VLMs.

@aclmeeting.bsky.social @genderbiasnlp.bsky.social

Frontiers | Dreams are more “predictable” than you think IntroductionA growing body of work has used machine learning and AI tools to analyse dream reports, and compare them to other textual content. Since these to...

👋 #sleeppeeps In this new work, we show how multiple LLMs appear to easily process #dreams, and reflect previously observed differences in reports (e.g., male-female, blind-sighted).

We also argue for caution linking dreams and bizarreness when using #LLMs

www.frontiersin.org/journals/sle...

Happy to share that our work on #conformalprediction for pulmonary nodule detection has been accepted for presentation at COPA 25 conference!

An epidemiological knowledge graph extracted from the World Health Organization’s Disease Outbreak News - Scientific Data Scientific Data - An epidemiological knowledge graph extracted from the World Health Organization’s Disease Outbreak News

www.nature.com/articles/s41...

Introducing PreDA (prefix-based dream annotation), a set of generative LLMs tuned to annotate dream reports for 6 Hall & Van De Castle features! The paper will be presented at LOD 25 conference later this year as a long contribution. Models available here huggingface.co/jrc-ai
#sleeppeeps #dream

Roel Hulsman, Valentin Comte, Lorenzo Bertolini, Tobias Wiesenthal, Antonio Puertas Gallardo, Mario Ceresa
Conformal Risk Control for Pulmonary Nodule Detection
https://arxiv.org/abs/2412.20167

💤 As I approach the end of my PostDoc at IMT Lucca, I’m looking for a new job!

🧠 Ideally: Research roles related to sleep/dream/consciousness/cognitive science.

🤖 Open to scientific writing & data analysis, too.

📍 Milan, nearby areas, or Switzerland.

DMs open 💃🏼

Would computational approaches like open source toolkit and/or trained models (e.g. LLMs) be of interest too?

Understanding conscious experiences during sleep Emerging research suggests that elements of awareness persist throughout various sleep stages, challenging conventional notions of unconsciousness during sleep. This dynamic and enigmatic interplay of...

Contribute to our Research Topic "Understanding conscious experiences during sleep"!
💤 🧠 💤

Deadlines:
Manuscript Summary Submission: 20.12.24
Manuscript Submission: 28.02.25

www.frontiersin.org/research-top...

#consciousexperiences #sleep #REMdreams #NREMdreams #DreamEngineering #Hypnagogia

thinking of calling this "The Illusion Illusion"

(more examples below)

My deep learning course at the University of Geneva is available on-line. 1000+ slides, ~20h of screen-casts. Full of examples in PyTorch.

fleuret.org/dlc/

And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)

fleuret.org/lbdl/

It's Sunday morning so taking a minute for a nerdy thread (on math, tokenizers and LLMs) of the work of our intern Garreth

By adding a few lines of code to the base Llama 3 tokenizer, he got a free boost in arithmetic performance 😮

[thread]

Long COVID: a clinical update Post-COVID-19 condition (also known as long COVID) is generally defined as symptoms persisting for 3 months or more after acute COVID-19. Long COVID can affect multiple organ systems and lead to sever...

Reproducing my thread from the Other Place on our Lancet paper 'Long COVID - a clinical update'. Coauthors Manoj Sivan, Alice Perlowski, Janko Nikolich. 1/

www.thelancet.com/journals/lan...?

E quindi come si sta da queste parti?

Work in progress -- suggestions for NLP-ers based in the EU/Europe & already on Bluesky very welcome!

go.bsky.app/NZDc31B

Heya, happy to be on the list if there is still a slot 🙂

Good question, but o don’t think so. I haven’t seen/found anyone yet.

Thanks, but it seems that the response is quite full of errors/hallucinations. In the paper of the first two datasets, there is no mention of existing age-based metadata (nor in the dataset download web page), while the third I donìt think it qualifies as histopath.

Hi alI quick dataset question:

I’m looking for a collection of histopathological images (no specific requirements wrt organ/disease/task) which includes metadata on participants’ (approximate) age.

Any suggestion?

A useful and very informed add to the debate would also be @lampinen.bsky.social 🙂

YouTube video by Manlio De Domenico The “great Xodus” and the growth of BlueSky

As a computational social scientist I could not miss to follow this beautiful creature: the BlueSky social network during its growth in the last hours!

In this video I show just a snapshot for about 210k users and their 500k-ish connections #academicsky

Zoom in the next post

youtu.be/JcmYZtWGnjQ

Nice to see so many new #sleeppeeps on #SleepSky! I’m just gonna leave this here again 😅

(btw the paper and DReAMy abstract have now been published!)

Papers at #EMNLP2024 #1

Statistical Uncertainty in Word Embeddings: GloVe-V

Neural models, from word vectors through transformers, use point estimate representations. They can have large variances, which often loom large in CSS applications.

Tue Nov 12 15:15-15:30 Flagler

New here? Interested in AI/ML? Check out these great starter packs!

AI: go.bsky.app/SipA7it
RL: go.bsky.app/3WPHcHg
Women in AI: go.bsky.app/LaGDpqg
NLP: go.bsky.app/SngwGeS
AI and news: go.bsky.app/5sFqVNS

You can also search all starter packs here: blueskydirectory.com/starter-pack...

Title: "Characterizing the Role of Similarity in the Property Inferences of Language Models" Authors: Juan Diego Rodriguez, Aaron Mueller, Kanishka Misra Left figure: "Given that dogs are daxable, is it true that corgis are daxable?" A language model could answer this either using taxonomic relations, illustrated by a taxonomy dog-corgi, dog-mutt, canine-wolf, etc., or by similarity relations (dogs are more similar to corgis than cats, wolves or shar peis). Right figure: illustration of the causal model (and an example intervention) for distributed alignment search (DAS), which we used to find a subspace in the network responsible for property inheritance behavior. The bottom nodes are "property", "premise concept (A)" and "conclusion concept (B)" , the middle nodes are "A has property P", "B is a kind of A", and the top node is "B has property P".

How do language models organize concepts and their properties? Do they use taxonomies to infer new properties, or infer based on concept similarities? Apparently, both!

🌟 New paper with my fantastic collaborators @amuuueller.bsky.social and @kanishka.bsky.social

GitHub - lorenzoscottb/DReAMy: DReAMy: a library for dream-reports annotation methods with python, NLP, and LLMs DReAMy: a library for dream-reports annotation methods with python, NLP, and LLMs - lorenzoscottb/DReAMy

Sad to miss the #neuroscience of #dreaming summers school and the opportunity to chat about #NLP and #LLMs for #dream research.

If you're interested, check out dreamy's tutorials on the Git page (also available on Google Colab) 😀

github.com/lorenzoscott...