Elliot Vincent
@elliotvincent.bsky.social
27🇫🇷 Researcher and AI Coordinator at @ignfrance.bsky.social. Interested in unsupervised learning, remote sensing, satellite imagery and scene understanding.
🏳️🌈 he/him
🏳️🌈 he/him
Reposted by Elliot Vincent
but also if that's really the angle they want to take, any trivial QA agent equipped with access to wikipedia satisfies a definition of AGI.
October 17, 2025 at 5:16 PM
but also if that's really the angle they want to take, any trivial QA agent equipped with access to wikipedia satisfies a definition of AGI.
Reposted by Elliot Vincent
@elliot-eu.bsky.social … can we have an ELLIOT starter pack?
October 11, 2025 at 9:26 AM
@elliot-eu.bsky.social … can we have an ELLIOT starter pack?
Reposted by Elliot Vincent
Démontant un à un les arguments avancés contre la taxe Zucman, l’économiste américain Joseph Stiglitz décrit l’égalité fiscale et la taxation des ultrariches comme des conditions de «la démocratie et de la solidarité sociale».
➡️ bit.ly/3KC9TnI
➡️ bit.ly/3KC9TnI
October 10, 2025 at 3:43 PM
Démontant un à un les arguments avancés contre la taxe Zucman, l’économiste américain Joseph Stiglitz décrit l’égalité fiscale et la taxation des ultrariches comme des conditions de «la démocratie et de la solidarité sociale».
➡️ bit.ly/3KC9TnI
➡️ bit.ly/3KC9TnI
Also about reviewing for #aaai: is it just me or is it painful to review without line numbering?
I know it’s not great to do things last minute, but I have to say, having read 4 papers and written their reviews, and not being able to submit them because openreview is down is really frustrating. Now the site is back but the review task is closed 🙃
#aaai
#openreview
#aaai
#openreview
September 26, 2025 at 2:26 PM
Also about reviewing for #aaai: is it just me or is it painful to review without line numbering?
Just discovered this existed and was set to On by default, you may want to opt out
September 26, 2025 at 7:16 AM
Just discovered this existed and was set to On by default, you may want to opt out
Reposted by Elliot Vincent
5,290 papers will be published at NeurIPS this year. If you could still get the proceedings in a book, it would be like 50,000 pages long. That’s 14.5 papers a day to read each and every day until the next NeurIPS
September 18, 2025 at 7:06 PM
5,290 papers will be published at NeurIPS this year. If you could still get the proceedings in a book, it would be like 50,000 pages long. That’s 14.5 papers a day to read each and every day until the next NeurIPS
Reposted by Elliot Vincent
Breaking news from #AI4EO symposium: cows confirmed as extraterrestrial beings. 👽 #CowsFromSpace
More information to come.
More information to come.
September 12, 2025 at 8:11 AM
Breaking news from #AI4EO symposium: cows confirmed as extraterrestrial beings. 👽 #CowsFromSpace
More information to come.
More information to come.
Reposted by Elliot Vincent
Hey you wonderful EO scientists. Why not nominate your amazing colleague or team or nominate yourself for the next ESA EO excellence award. Nominations have just opened and close 10th Dec. Research cash prize for lucky winners:
www.eoxcellence.com
www.eoxcellence.com
September 12, 2025 at 11:57 AM
Hey you wonderful EO scientists. Why not nominate your amazing colleague or team or nominate yourself for the next ESA EO excellence award. Nominations have just opened and close 10th Dec. Research cash prize for lucky winners:
www.eoxcellence.com
www.eoxcellence.com
I know it’s not great to do things last minute, but I have to say, having read 4 papers and written their reviews, and not being able to submit them because openreview is down is really frustrating. Now the site is back but the review task is closed 🙃
#aaai
#openreview
#aaai
#openreview
September 2, 2025 at 1:47 PM
I know it’s not great to do things last minute, but I have to say, having read 4 papers and written their reviews, and not being able to submit them because openreview is down is really frustrating. Now the site is back but the review task is closed 🙃
#aaai
#openreview
#aaai
#openreview
Reposted by Elliot Vincent
I wrote an op-ed on the world-class STEM research ecosystem in the United States, and how this ecosystem is now under attack on multiple fronts by the current administration: newsletter.ofthebrave.org/p/im-an-awar...
I’m an award-winning mathematician. Trump just cut my funding.
The “Mozart of Math” tried to stay out of politics. Then it came for his research.
newsletter.ofthebrave.org
August 18, 2025 at 3:45 PM
I wrote an op-ed on the world-class STEM research ecosystem in the United States, and how this ecosystem is now under attack on multiple fronts by the current administration: newsletter.ofthebrave.org/p/im-an-awar...
Reposted by Elliot Vincent
🚀 DinoV3 just became the new go-to backbone for geoloc!
It outperforms CLIP-like models (SigLip2, finetuned StreetCLIP)… and that’s shocking 🤯
Why? CLIP models have an innate advantage — they literally learn place names + images. DinoV3 doesn’t.
It outperforms CLIP-like models (SigLip2, finetuned StreetCLIP)… and that’s shocking 🤯
Why? CLIP models have an innate advantage — they literally learn place names + images. DinoV3 doesn’t.
August 18, 2025 at 3:14 PM
🚀 DinoV3 just became the new go-to backbone for geoloc!
It outperforms CLIP-like models (SigLip2, finetuned StreetCLIP)… and that’s shocking 🤯
Why? CLIP models have an innate advantage — they literally learn place names + images. DinoV3 doesn’t.
It outperforms CLIP-like models (SigLip2, finetuned StreetCLIP)… and that’s shocking 🤯
Why? CLIP models have an innate advantage — they literally learn place names + images. DinoV3 doesn’t.
Reposted by Elliot Vincent
Reposted by Elliot Vincent
Academia is basically a collection of people who got lucky early on and mistook it for genius. doi.org/10.1073/pnas...
August 4, 2025 at 11:26 AM
Academia is basically a collection of people who got lucky early on and mistook it for genius. doi.org/10.1073/pnas...
Reposted by Elliot Vincent
I had the privilege to be invited to speak about our work "Around the World in 80 Timesteps" at the French Podcast Underscore! If you speak french, i highly recommend it they did a great job with the montage!
If you want to learn more nicolas-dufour.github.io/plonk
www.youtube.com/watch?v=s5oH...
If you want to learn more nicolas-dufour.github.io/plonk
www.youtube.com/watch?v=s5oH...
Il a conçu la première IA d’OSINT (terrifiant… et génial)
YouTube video by Underscore_
www.youtube.com
July 31, 2025 at 4:43 PM
I had the privilege to be invited to speak about our work "Around the World in 80 Timesteps" at the French Podcast Underscore! If you speak french, i highly recommend it they did a great job with the montage!
If you want to learn more nicolas-dufour.github.io/plonk
www.youtube.com/watch?v=s5oH...
If you want to learn more nicolas-dufour.github.io/plonk
www.youtube.com/watch?v=s5oH...
Reposted by Elliot Vincent
Reposted by Elliot Vincent
Bluesky doesn't penalize you for using links, tags, using key words like Patreon or commissions, and doesn't have a central algorithm on your following feed, which means all the power rests on you to help your favourite creators be seen!
July 30, 2025 at 6:05 PM
Bluesky doesn't penalize you for using links, tags, using key words like Patreon or commissions, and doesn't have a central algorithm on your following feed, which means all the power rests on you to help your favourite creators be seen!
Reposted by Elliot Vincent
💥 2 MILLIONS DE SIGNATURES 💥
Vous l’avez fait ! La pétition contre la loi Duplomb vient de franchir le cap des 2 millions de signatures.
Cette réaction citoyenne est inédite. Pour les pro-Duplomb, elle serait le fruit d’une «séquence de désinformation». On détricote le vrai du faux juste ici👇
Vous l’avez fait ! La pétition contre la loi Duplomb vient de franchir le cap des 2 millions de signatures.
Cette réaction citoyenne est inédite. Pour les pro-Duplomb, elle serait le fruit d’une «séquence de désinformation». On détricote le vrai du faux juste ici👇
Deux millions de signatures contre la loi Duplomb : les partisans du texte controversé dénoncent une «désinformation» écolo
Ils ne manquent pas Duplomb. La pétition contre cette loi décriée a dépassé les deux millions de signatures, ce lundi 28 juillet. Pour les pro-Duplomb, cette réaction citoyenne serait le fruit d’une «...
vert.eco
July 28, 2025 at 6:31 AM
💥 2 MILLIONS DE SIGNATURES 💥
Vous l’avez fait ! La pétition contre la loi Duplomb vient de franchir le cap des 2 millions de signatures.
Cette réaction citoyenne est inédite. Pour les pro-Duplomb, elle serait le fruit d’une «séquence de désinformation». On détricote le vrai du faux juste ici👇
Vous l’avez fait ! La pétition contre la loi Duplomb vient de franchir le cap des 2 millions de signatures.
Cette réaction citoyenne est inédite. Pour les pro-Duplomb, elle serait le fruit d’une «séquence de désinformation». On détricote le vrai du faux juste ici👇
Reposted by Elliot Vincent
Loi Duplomb : une pétition avec 500 000 signatures rouvre la voie à une discussion à l'Assemblée
Loi Duplomb : une pétition avec 500 000 signatures rouvre la voie à une discussion à l'Assemblée
Une pétition demandant l'abrogation de la loi Duplomb et sa mesure très contestée de réintroduction d'un pesticide a réuni samedi 500 000 signatures sur le site de l'Assemblée nationale. Du jamais-vu pour une initiative de ce type, ce qui rouvre la voie à une discussion sur le bien-fondé de cette loi mais pas à son réexamen.
f24.my
July 19, 2025 at 2:24 PM
Loi Duplomb : une pétition avec 500 000 signatures rouvre la voie à une discussion à l'Assemblée
Reposted by Elliot Vincent
🌍 Would be better if authors were allowed to present at officially endorsed hubs or online instead — ensuring fair and equitable participation, and ultimately creating a better NeurIPS for everyone.
Interesting in knowing more: read our full opinion piece ⬇️. docs.google.com/document/d/1...
4/4
Interesting in knowing more: read our full opinion piece ⬇️. docs.google.com/document/d/1...
4/4
Towards decentralized ML conferences
Towards decentralized ML conferences Pierre Marion July 2025 Abstract: Large ML conferences play a central role in our community, by providing a rigorous review process and a venue for meaningful di...
docs.google.com
July 17, 2025 at 8:49 AM
🌍 Would be better if authors were allowed to present at officially endorsed hubs or online instead — ensuring fair and equitable participation, and ultimately creating a better NeurIPS for everyone.
Interesting in knowing more: read our full opinion piece ⬇️. docs.google.com/document/d/1...
4/4
Interesting in knowing more: read our full opinion piece ⬇️. docs.google.com/document/d/1...
4/4
Reposted by Elliot Vincent
📢 Present your NeurIPS paper in Europe!
Join EurIPS 2025 + ELLIS UnConference in Copenhagen for in-person talks, posters, workshops and more. Registration opens soon; save the date:
📅 Dec 2–7, 2025
📍 Copenhagen 🇩🇰
🔗eurips.cc
#EurIPS
@euripsconf.bsky.social
Join EurIPS 2025 + ELLIS UnConference in Copenhagen for in-person talks, posters, workshops and more. Registration opens soon; save the date:
📅 Dec 2–7, 2025
📍 Copenhagen 🇩🇰
🔗eurips.cc
#EurIPS
@euripsconf.bsky.social
July 16, 2025 at 11:00 PM
📢 Present your NeurIPS paper in Europe!
Join EurIPS 2025 + ELLIS UnConference in Copenhagen for in-person talks, posters, workshops and more. Registration opens soon; save the date:
📅 Dec 2–7, 2025
📍 Copenhagen 🇩🇰
🔗eurips.cc
#EurIPS
@euripsconf.bsky.social
Join EurIPS 2025 + ELLIS UnConference in Copenhagen for in-person talks, posters, workshops and more. Registration opens soon; save the date:
📅 Dec 2–7, 2025
📍 Copenhagen 🇩🇰
🔗eurips.cc
#EurIPS
@euripsconf.bsky.social
Reposted by Elliot Vincent
Talking politics at work in 2025 is like "Now I'm not trying to get all political, but I think Hitler is bad."
July 11, 2025 at 2:38 AM
Talking politics at work in 2025 is like "Now I'm not trying to get all political, but I think Hitler is bad."
Reposted by Elliot Vincent
@cnrsecologie.bsky.social regrette profondément l'adoption de cette loi à la vision court-termiste & ses conséquences graves sur l’environnement, qui méprise santé & bien-être de la population & le rôle des espèces sauvages dans la prod. agricole. La communauté scientifique n'a pas été entendue.
Le Parlement a définitivement adopté mardi la proposition de loi agricole Duplomb-Menonville, comprenant entre autres une mesure décriée de réintroduction sous conditions d'un pesticide néonicotinoïde, et présentée comme une réponse à la colère agricole de 2024.
July 9, 2025 at 8:38 AM
@cnrsecologie.bsky.social regrette profondément l'adoption de cette loi à la vision court-termiste & ses conséquences graves sur l’environnement, qui méprise santé & bien-être de la population & le rôle des espèces sauvages dans la prod. agricole. La communauté scientifique n'a pas été entendue.
Reposted by Elliot Vincent
Je viens de le remarquer mais des bots reprennent ici les comptes officiels des RER (très pratique) :
@rer-a.bsky.social
@rer-b.bsky.social
@rer-c.bsky.social
@rer-d.bsky.social
@rer-e.bsky.social
@rer-a.bsky.social
@rer-b.bsky.social
@rer-c.bsky.social
@rer-d.bsky.social
@rer-e.bsky.social
July 6, 2025 at 7:40 PM
Je viens de le remarquer mais des bots reprennent ici les comptes officiels des RER (très pratique) :
@rer-a.bsky.social
@rer-b.bsky.social
@rer-c.bsky.social
@rer-d.bsky.social
@rer-e.bsky.social
@rer-a.bsky.social
@rer-b.bsky.social
@rer-c.bsky.social
@rer-d.bsky.social
@rer-e.bsky.social
Reposted by Elliot Vincent
Pour les collègues francophones, vous saviez que le FID était tout cassé ? Moi non plus. Pourtant, si on s'y prend bien, on peut calculer le FID avec moins de 1000 images.
J'en parlerai au GRETSI fin août : hal.science/hal-05142942 👀
J'en parlerai au GRETSI fin août : hal.science/hal-05142942 👀
Évaluation des générateurs d'images à partir de peu d'exemples : calculer le FID avec 10 fois moins d'images, c'est possible
La distance Inception de Fréchet (Fréchet Inception Distance ou FID) est une métrique standard pour l'évaluation des modèles génératifs images. Construite sur la distance de Wasserstein, le FID mesure...
hal.science
July 9, 2025 at 2:12 PM
Pour les collègues francophones, vous saviez que le FID était tout cassé ? Moi non plus. Pourtant, si on s'y prend bien, on peut calculer le FID avec moins de 1000 images.
J'en parlerai au GRETSI fin août : hal.science/hal-05142942 👀
J'en parlerai au GRETSI fin août : hal.science/hal-05142942 👀