François Fleuret
@francois.fleuret.org
Research Scientist Meta/FAIR, Prof. University of Geneva, co-founder Neural Concept SA. I like reality.
https://fleuret.org
https://fleuret.org
Pinned
François Fleuret
@francois.fleuret.org
· Nov 26
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/
fleuret.org/dlc/
And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!)
fleuret.org/lbdl/
I asked "on the other platform" what were the most important improvements to the original 2017 transformer.
That was quite popular and here is a synthesis of the responses:
That was quite popular and here is a synthesis of the responses:
April 28, 2025 at 6:47 AM
I asked "on the other platform" what were the most important improvements to the original 2017 transformer.
That was quite popular and here is a synthesis of the responses:
That was quite popular and here is a synthesis of the responses:
"You are in Paris, enjoy the city, stop obsessing with AI"
Paris:
Paris:
April 1, 2025 at 4:18 PM
"You are in Paris, enjoy the city, stop obsessing with AI"
Paris:
Paris:
Maybe the wall was the friends we made during that journey Ted.
The year is 2435. Human beings — now sentient spheres of glowing gas — finally understand why matter exists. Our knowledge of the world is complete, and can go no farther!
One the telepresence screen, a simulation of a 21c internet pundit pops up: "Told you deep learning was hitting a wall!"
One the telepresence screen, a simulation of a 21c internet pundit pops up: "Told you deep learning was hitting a wall!"
February 28, 2025 at 6:53 AM
Maybe the wall was the friends we made during that journey Ted.
Reposted by François Fleuret
What is the true depth of an LLM?
Together with @danielepal.bsky.social , @matpagliardini.bsky.social, M. Jaggi and @francois.fleuret.org we show that LLMs have a smaller effective depth that can be exploited to increase inference speeds on multi-GPU settings!
arxiv.org/abs/2502.02790
(1/N)
Together with @danielepal.bsky.social , @matpagliardini.bsky.social, M. Jaggi and @francois.fleuret.org we show that LLMs have a smaller effective depth that can be exploited to increase inference speeds on multi-GPU settings!
arxiv.org/abs/2502.02790
(1/N)
February 14, 2025 at 4:17 PM
What is the true depth of an LLM?
Together with @danielepal.bsky.social , @matpagliardini.bsky.social, M. Jaggi and @francois.fleuret.org we show that LLMs have a smaller effective depth that can be exploited to increase inference speeds on multi-GPU settings!
arxiv.org/abs/2502.02790
(1/N)
Together with @danielepal.bsky.social , @matpagliardini.bsky.social, M. Jaggi and @francois.fleuret.org we show that LLMs have a smaller effective depth that can be exploited to increase inference speeds on multi-GPU settings!
arxiv.org/abs/2502.02790
(1/N)
J'étais l'invité du journal de 19h30 sur la @radiotelesuisse.bsky.social ce soir pour parler d'Intelligence Artificielle.
www.rts.ch/play/tv/19h3...
www.rts.ch/play/tv/19h3...
19h30 - Play RTS
Play RTS vous permet de visionner ou d'écouter de nombreuses émissions tv ou radio, quand et aussi souvent que vous le souhaitez.
www.rts.ch
February 11, 2025 at 11:08 PM
J'étais l'invité du journal de 19h30 sur la @radiotelesuisse.bsky.social ce soir pour parler d'Intelligence Artificielle.
www.rts.ch/play/tv/19h3...
www.rts.ch/play/tv/19h3...
It is hard to overstate how cool and powerful is flex attention. @chhillee.bsky.social
pytorch.org/blog/flexatten…
TL;DR: it is an implementation of the attention operator in pytorch that allows in particular to efficiently "carve" the attention matrix.
1/3
pytorch.org/blog/flexatten…
TL;DR: it is an implementation of the attention operator in pytorch that allows in particular to efficiently "carve" the attention matrix.
1/3
https://pytorch.org/blog/flexatten…
February 6, 2025 at 12:23 AM
It is hard to overstate how cool and powerful is flex attention. @chhillee.bsky.social
pytorch.org/blog/flexatten…
TL;DR: it is an implementation of the attention operator in pytorch that allows in particular to efficiently "carve" the attention matrix.
1/3
pytorch.org/blog/flexatten…
TL;DR: it is an implementation of the attention operator in pytorch that allows in particular to efficiently "carve" the attention matrix.
1/3
Reposted by François Fleuret
Finally got this beautiful piece from @francois.fleuret.org
January 10, 2025 at 10:44 AM
Finally got this beautiful piece from @francois.fleuret.org
Happy new year you all!
2025 is certainly full of promise.
2025 is certainly full of promise.
December 31, 2024 at 11:55 PM
Happy new year you all!
2025 is certainly full of promise.
2025 is certainly full of promise.
Happy Christmas you all!
December 25, 2024 at 12:32 AM
Happy Christmas you all!
Whatever you say about the whole field of AI: It's not boring.
December 22, 2024 at 8:57 AM
Whatever you say about the whole field of AI: It's not boring.
Oh boy, GTA 6 has to be good.
And Half Life 3.
And Half Life 3.
December 19, 2024 at 3:40 PM
Oh boy, GTA 6 has to be good.
And Half Life 3.
And Half Life 3.
This is very great.
Alongside our paper, we also recorded a roundtable video featuring four of the paper’s authors discussing the results and their implications in detail:
Alignment faking in large language models
YouTube video by Anthropic
www.youtube.com
December 19, 2024 at 8:29 AM
This is very great.
You young post-2012 researchers, you realise that us the old people are *amazed* by what we witness?
December 18, 2024 at 5:54 AM
You young post-2012 researchers, you realise that us the old people are *amazed* by what we witness?
Something that annoys me much is when people (generally in political / "societal" discussions) equate the validity of a statement with the validity of *what "people" may understand* and generally then fix a true statement they consider false with a false statement they consider true.
1/2
1/2
December 18, 2024 at 5:37 AM
Something that annoys me much is when people (generally in political / "societal" discussions) equate the validity of a statement with the validity of *what "people" may understand* and generally then fix a true statement they consider false with a false statement they consider true.
1/2
1/2
Reposted by François Fleuret
hello bluesky! we have a new preprint on solvation free energies:
tl;dr: We define an interpolating density by its sampling process, and learn the corresponding equilibrium potential with score matching. arxiv.org/abs/2410.15815
with @francois.fleuret.org and @tbereau.bsky.social
(1/n)
tl;dr: We define an interpolating density by its sampling process, and learn the corresponding equilibrium potential with score matching. arxiv.org/abs/2410.15815
with @francois.fleuret.org and @tbereau.bsky.social
(1/n)
December 17, 2024 at 12:32 PM
hello bluesky! we have a new preprint on solvation free energies:
tl;dr: We define an interpolating density by its sampling process, and learn the corresponding equilibrium potential with score matching. arxiv.org/abs/2410.15815
with @francois.fleuret.org and @tbereau.bsky.social
(1/n)
tl;dr: We define an interpolating density by its sampling process, and learn the corresponding equilibrium potential with score matching. arxiv.org/abs/2410.15815
with @francois.fleuret.org and @tbereau.bsky.social
(1/n)
"The expletive error rate is the number of times you call your algorithm a 'dense mf' when looking at one run"
December 17, 2024 at 11:09 AM
"The expletive error rate is the number of times you call your algorithm a 'dense mf' when looking at one run"
General structure of a paper:
- general ideas
- general case
- general case
- general case
- what we actually do
how it should be:
- what we actually do
- why we think it's great as one method of a general class
- how we got there
- how we got there
- how we got there
- general ideas
- general case
- general case
- general case
- what we actually do
how it should be:
- what we actually do
- why we think it's great as one method of a general class
- how we got there
- how we got there
- how we got there
December 17, 2024 at 10:55 AM
General structure of a paper:
- general ideas
- general case
- general case
- general case
- what we actually do
how it should be:
- what we actually do
- why we think it's great as one method of a general class
- how we got there
- how we got there
- how we got there
- general ideas
- general case
- general case
- general case
- what we actually do
how it should be:
- what we actually do
- why we think it's great as one method of a general class
- how we got there
- how we got there
- how we got there
Today's mood "wow GPTs are dumb"
December 17, 2024 at 10:44 AM
Today's mood "wow GPTs are dumb"
One of the goosebumpsiest short movie ever (where this quote comes from). We will have this sound snippet played at the ceremony when we will leave earth for the stars.
www.youtube.com/watch?v=YH3c...
www.youtube.com/watch?v=YH3c...
December 17, 2024 at 10:15 AM
One of the goosebumpsiest short movie ever (where this quote comes from). We will have this sound snippet played at the ceremony when we will leave earth for the stars.
www.youtube.com/watch?v=YH3c...
www.youtube.com/watch?v=YH3c...
"We invest far-off places with a certain romance. This appeal, I suspect, has been meticulously crafted by natural selection as an essential element in our survival."
December 17, 2024 at 10:07 AM
"We invest far-off places with a certain romance. This appeal, I suspect, has been meticulously crafted by natural selection as an essential element in our survival."
AlphaGo - The Movie | Full award-winning documentary
YouTube video by Google DeepMind
m.youtube.com
December 16, 2024 at 3:39 AM