
Particle Physicist at @IJClab @ATLASexperiment at @CERN. #AI and physics. I have left this officially na$i platform, you know where, same handle.
Joined March 2011
- Tweets2K
- Following1.2K
- Followers983
- Likes4.4K
Pinned Tweet
Is uncertainties handling the ultimate roadblock for a wider adoption of AI in science ?
AI and the Uncertainty Challenge in Fundamental Physics, 27-Nov-1-Dec
with sessions on Uncertainty Quantification, Simulation Based Inference and more, indico.in2p3.fr/event/30589/
Je viens de faire mon #eXit! L’exode de X est massif. Ne perdez pas un seul de vos abonnés. Grâce à #HelloQuitX j’ai inscrit 2265 nouveaux passagers pour partir vers #BlueSky & #Mastodon. Retrouvez automatiquement vos communautés le #20Janvier via app.helloquitx.com !
Just created the “boomers @ Neurips” group in #neurips Whova app. I had to.
David Rousseau retweeted
Career update 📢—I’m on the job market for faculty positions! My work focuses on advancing particle physics 💥 and astrophysics 🌌 by creating innovative, robust statistical methods with AI 🤖. If you know of any relevant openings, please reach out!
David Rousseau retweeted
Excited to share the first results using Neural Simulation-Based Inference (NSBI) techniques applied to @ATLASexperiment data! We measure the elusive off-shell Higgs boson with 3.1x better observation sensitivity than standard (histogram) analysis techniques! A thread: (1/N)
David Rousseau retweeted
I'm happy and proud to present the results of the #CaloChallenge!
arxiv.org/abs/2410.21611
A huge thank you to everyone who contributed to this effort in the past year(s)!
1/12
David Rousseau retweeted
But making it happen in an actual experiment at the LHC took a long time. Check out the thread by @Aishik_Ghosh_ about his 6-year journey. And shoutout to @dhpmrou for being an awesome advisor that saw the potential for these techniques early on
xitter.cf/glouppe/status/1851168…
They made it! @Aishik_Ghosh_, Jay Sandesara, @dhpmrou and Rafael CLdS pushed through the first neural simulation-based inference analysis @ATLASexperiment. Huge congratulations to everyone involved!
David Rousseau retweeted
A thread on quantum interference and the 6 year adventure it took me through
In experimental particle physics we sift out events from signal (eg. Higgs) processes vs the orders of magnitude more events from background processes…
.@brucedenby is the author of the very first paper on AI and Particle Physics back in 1988, when he was a post-doc in Orsay ( @IJCLab @cnrs_in2p3) working in the @CERN LEP DELPHI experiment.
He was almost fired for this!
#HEPML @dorigo @KyleCranmer @GregorKasieczka
He told us his story in the Learning to Discover workshop I helped organised in April 2022 at @pascalinstitut sponsored by @dataia
indico.ijclab.in2p3.fr/event… (search for "Denby" for the talk slides and video )
AI and physics is not a solved problem ! Check out the #NeuriPS 2024 Fair Universe HIggsML Uncertainty challenge to develop AI algorithms measuring the physics properties of elementary particles with imperfect simulators, with controlled uncertainties.
fair-universe.lbl.gov
I was today's year old when I realised the Cerenkov light cone was *behind*, not *in front of*, the particle travelling faster than light in the material (from left to right here) .
Completely obvious of course. Shame, shame.
indico.in2p3.fr/event/33412/…
... the link to the original kaggle forum post for reference: kaggle.com/competitions/higg…
David Rousseau retweeted
Looking for a good example of use-inspired research where a physics problem fosters a highly impactful general-purpose method? I’m not talking about MCMC or HMC, I’m talking about @XGBoostProject
Tagging the #hepml community : @KyleCranmer @ClaudiusKrause @DanielWhiteson @SergeiGleyzer @BPNachman @Stefanog @HEPfeickert @glouppe @Giles_C_Strong @DaniloJRezende @SaltyBurger @vlimant @lukasheinrich_ @michaelkagan @dorigo @pietrovischia














