Pinned Tweet
Our new paper is finally out! Read the thread by @hippopedoid for some highlights, and don't miss the interactive visualization by @benmschmidt ✨
static.nomic.ai/pubmed.html
Work with @CellTypist and others.
Really excited to present new work by @ritagonmar: we visualized the entire PubMed library, 21 million biomedical and life science papers, and learned a lot about --
THE LANDSCAPE OF BIOMEDICAL RESEARCH
biorxiv.org/content/10.1101/…
Joint work with @CellTypist and @benmschmidt. 1/n
Rita González Márquez @ritagonmar.bsky.social retweeted
We updated our ICLR dataset (see github.com/berenslab/iclr-da…) with blind 2025 submissions to @iclr_conf. Over 10k submissions this year.
I like how this embedding (SBERT + tSNE) shows which ML areas are old-school and which ones are currently booming.
Rita González Márquez @ritagonmar.bsky.social retweeted
Researching #AI #ComputerVision #MachineLearning #Robotics #HCI? Join our elite doctoral program - a partnership with MPI-IS, @uni_stuttgart & @uni_tue! Applications accepted until Nov 15, 2024 at imprs.is.mpg.de/application
Rita González Márquez @ritagonmar.bsky.social retweeted
So she was first author on his key paper - in fact she's on most of his key papers, yet she didn't get the Nobel Prize, in spite of the fact they could have given 3 awards, not just 2. Explain to me how STEM is not institutionally sexist.
Congratulations to our 2024 medicine laureate Victor Ambros ✨
This morning he celebrated the news of his prize with his colleague and wife Rosalind Lee, who was also the first author on the 1993 'Cell' paper cited by the Nobel Committee.
#NobelPrize
Rita González Márquez @ritagonmar.bsky.social retweeted
Researching #AI #MachineLearning #Robotics #HCI? Join our elite doctoral program - a partnership with MPI-IS, @uni_stuttgart & @uni_tue! Applications accepted until Nov 15, 2024 at imprs.is.mpg.de/application
Rita González Márquez @ritagonmar.bsky.social retweeted
Excited to be in Vienna at ICML to present my work with @samgreydanus on scaling _down_ deep learning with MNIST-1D!
We show how one can study serious deep learning with n=5000 and d=40.
Paper: openreview.net/forum?id=n9pr…
Code: github.com/greydanus/mnist1d…
Find me at poster session 1.
Rita González Márquez @ritagonmar.bsky.social retweeted
Interested in reliable parameter inference for ODEs with probabilistic solvers?
I am excited to present my latest work w. @nathanaelbosch, @deismic_, @KadhimKyra, @jakhmack, @PhilippHennig5, @CellTypist, @ml4science at #ICML2024 this year. 1/n
Rita González Márquez @ritagonmar.bsky.social retweeted
LLMs are not just a faster kind of typewriter, they can introduce subtle mistakes and biases (as we wrote in our discussion). That can work in some scenarios, but may be a problem in others. Rita made that point nicely over here:
xitter.cf/ritagonmar/status/1804…
Rita González Márquez @ritagonmar.bsky.social retweeted
Fantastic work, estimating both baseline and departures from it - "excess mortality" like framework in an unexpected field! This makes it much less reliant on assumptions to detect LLMs.
Dmitry is absolutely one of the best in the growing literature of "Statistical-Forensics".
How many academic papers are written with the help of ChatGPT? To answer this question, we analyzed 14mln PubMed abstracts from 2010 to 2024 and looked for excess words:
** Delving into ChatGPT usage in academic writing through excess vocabulary **
arxiv.org/abs/2406.07016
1/11
Rita González Márquez @ritagonmar.bsky.social retweeted
What an estimate: "At least 10% of 2024 PubMed abstracts were processed with LLMs".
How many academic papers are written with the help of ChatGPT? To answer this question, we analyzed 14mln PubMed abstracts from 2010 to 2024 and looked for excess words:
** Delving into ChatGPT usage in academic writing through excess vocabulary **
arxiv.org/abs/2406.07016
1/11
Rita González Márquez @ritagonmar.bsky.social retweeted
@hippopedoid pointed to an important caveat when it comes to comparing countries: all of our estimates are lower bounds, so 3% GPT style words in UK does not exclude that true GPT usage there is higher but better disguised.
xitter.cf/hippopedoid/status/180…
Rita González Márquez @ritagonmar.bsky.social retweeted
Anyway, if you want to meticulously delve into our intricate analysis, here are the links:
Code: github.com/berenslab/chatgpt…
Paper: arxiv.org/abs/2406.07016
Big thanks to collaborators @ritagonmar, @JanLause, and Agnes Horvat.
Stay tuned for updates with more 2024 data!
11/11
Rita González Márquez @ritagonmar.bsky.social retweeted
How many scientists use #ChatGPT to help with their writing?🤖📝
We find #LLM style words in 10% of all 2024 PubMed abstracts, and up to 30% in some subfields. And that’s just lower bounds on the actual usage!
Read more in our paper or Dmitrys thread:
arxiv.org/abs/2406.07016
How many academic papers are written with the help of ChatGPT? To answer this question, we analyzed 14mln PubMed abstracts from 2010 to 2024 and looked for excess words:
** Delving into ChatGPT usage in academic writing through excess vocabulary **
arxiv.org/abs/2406.07016
1/11
Rita González Márquez @ritagonmar.bsky.social retweeted
How many academic papers are written with the help of ChatGPT? To answer this question, we analyzed 14mln PubMed abstracts from 2010 to 2024 and looked for excess words:
** Delving into ChatGPT usage in academic writing through excess vocabulary **
arxiv.org/abs/2406.07016
1/11
Our paper is featured in the cover of the Patterns' issue this month. How cool is the illustration by Franz Stämmele (staemmele.com/)? 🤩
Our June issue is now live!
cell.com/patterns/issue?pii=…
On the cover this month, researchers from the Tübingen AI Center use a large language model to explore and visualization the entirety of the biomedical literature in PubMed.
Full paper: cell.com/patterns/fulltext/S…
ALT Every year, over 1.5 million scientific articles are published in the fields of biomedicine and life sciences. This vast amount of publications makes it challenging to keep up with the literature and have an overview of new developments in the field. In this issue, González-Márquez et al. introduce a map to help navigate this complex landscape. The cover image shows a woman studying a map while pondering a landscape consisting of mountains of documents. Image credit: Franz-Georg Staemmele.
Rita González Márquez @ritagonmar.bsky.social retweeted
Our June issue is now live!
cell.com/patterns/issue?pii=…
On the cover this month, researchers from the Tübingen AI Center use a large language model to explore and visualization the entirety of the biomedical literature in PubMed.
Full paper: cell.com/patterns/fulltext/S…
ALT Every year, over 1.5 million scientific articles are published in the fields of biomedicine and life sciences. This vast amount of publications makes it challenging to keep up with the literature and have an overview of new developments in the field. In this issue, González-Márquez et al. introduce a map to help navigate this complex landscape. The cover image shows a woman studying a map while pondering a landscape consisting of mountains of documents. Image credit: Franz-Georg Staemmele.
Back from Vienna after a great week at ICLR. It was a good mix of inspiring science, meeting cool people, fun rollercoaster rides, and a beautiful city. Couldn't be more grateful!
Rita González Márquez @ritagonmar.bsky.social retweeted
Great to see the openTSNE paper finally published! Pavlin's openTSNE is by far the best t-SNE implementation out there. Very fast, very easy to use, but also very flexible and easy to extend and build upon. We use it in all our t-SNE-related projects. Great job @pavlinpolicar!
I am thrilled to announce that our paper on openTSNE is finally out in JSS. It's been nearly five years since we put up our preprint on bioarXiv, and much has changed since then. Here's the link to the paper: jstatsoft.org/article/view/v…
Rita González Márquez @ritagonmar.bsky.social retweeted
Finally, we found the authors with the most submissions (170 submissions in 8 years means >20 submissions per year btw...), highlighted their submissions in 2D, and saw some 🦊 and some 🦔, according to the famous classification by Isaiah Berlin: en.wikipedia.org/wiki/The_He…. 6/n
Rita González Márquez @ritagonmar.bsky.social retweeted
Links:
github.com/berenslab/iclr-da…
openreview.net/forum?id=2OOb…
arxiv.org/abs/2404.08403
If you are at @iclr_conf, go talk to @ritagonmar, she will present this work at the DMLR worhshop on Saturday. 7/7
Rita González Márquez @ritagonmar.bsky.social retweeted
Learn more about representation learning @iclr_conf with @ritagonmar and a dataset of ICLR papers! 📜🗺️⤵️
Here's some fun reading for all of you traveling to @iclr_conf today. A DMLR workshop paper by @ritagonmar with the title I'm very proud of: "Learning representations of learning representations" 😎 In it, we present and analyze the ICLR dataset. 1/n
github.com/berenslab/iclr-da…
