
PhD student in ML for atomistic simulations @lab_COSMO @EPFL ๐จ๐ญ | Previously: @3dsBIOVIA @isisneutronmuon @BristolUni ๐ฌ๐ง @UW ๐บ๐ธ
Lausanne, Switzerland
Joined July 2021
- Tweets42
- Following399
- Followers142
- Likes106
Joseph W. Abbott retweeted
Congratulations to John H. Hopfield and Geoffrey Hinton for winning the 2024 Physics Nobel Prize - and to MARVEL's @gppcarleo whose 2017 Science paper, on ANNs and the quantum many-body problem, is cited in the @NobelPrize scientific backgrounder ๐nobelprize.org/uploads/2024/โฆ
Joseph W. Abbott retweeted
๐จ chemiscope.org 0.7 is out ๐จ
thanks to brilliant @nccr_marvel #inspirepotentials fellow ๐บ๐ฆ Sofiia Chorna ๐๐๐ you can now include chemiscope viewers in documentation pages, using sphinx and/or sphinx-gallery. The possibilities are endless!
Joseph W. Abbott retweeted
I am presenting our paper about smooth symmetrization arxiv.org/abs/2305.19302 this Monday, 22.04, 11 am ET, at the LoGaG reading group; see portal.valencelabs.com/logg for the Zoom link.
Should we use equivariant or unconstrained GNNs?๐ค
Will be happy to see you there to discuss๐ค
Come and see @HowWeiBin, Hanna, @Divya_Suman14 and I talk about ML for electronic structure at #DPG this morning!
Replying to @lab_COSMO
๐๏ธ Thursday / 1 :
โ๏ธ 11:00 - Surrogate models for the electron density and related scalar fields - Joseph Abbott (O 82.3)
โ๏ธ 11:15 - Conventional definitions of the absolute energy reference and machine learning - Wei Bin How (O 82.4)
Joseph W. Abbott retweeted
COSMO @ DPG tomorrow: session O82 is an all-star show with Joe, Wei Bin, Hanna & Divvya!
Replying to @lab_COSMO
๐๏ธ Thursday / 1 :
โ๏ธ 11:00 - Surrogate models for the electron density and related scalar fields - Joseph Abbott (O 82.3)
โ๏ธ 11:15 - Conventional definitions of the absolute energy reference and machine learning - Wei Bin How (O 82.4)
Joseph W. Abbott retweeted
Can you decode our special message? ๐ต๏ธโโ๏ธโจ
P. S. HAPPY INTERNATIONAL WOMEN'S DAY to all the inspiring women out there! ๐ฉโ๐ฌ๐ฉโ๐จ๐ฉโ๐ณ๐ฉ๐ฟโ๐ป๐ฉโ๐พ๐ฉโ๐๐ฉ๐ปโโ๏ธ๐ฉโโ๏ธ๐๐พ๐ฉโโ๏ธ๐ฉโ๐ญ๐ง๐ผ๐ฆธ๐ฝโโ๏ธ๐งโโ๏ธ๐คธโโ๏ธ๐ท๐พโโ๏ธ๐ฉ๐ผโ๐ค๐ฉโ๐๐๏ธโโ๏ธ๐ฎโโ๏ธ๐๐พโโ๏ธ๐ฉโ๐ซ๐ฉ๐ฝโ๐ผ
P. P. S. Don't *see* the answer?๐ DW, our prof struggled too ๐@pschwllr
#EPFL #EPFLChem #InternationalWomensDay
Joseph W. Abbott retweeted
๐ข Shoutout for ๐จโ๐ @linuskellner's first #preprint, discussing DPOSE: a simple, accurate and inexpensive uncertainty quantification scheme for #MachineLearning potentials (and more) arxiv.org/html/2402.16621v1
Joseph W. Abbott retweeted
Great collaboration indeed, and so good to see it out in pretty format. If you want a teaser before you read the paper, have a look at the summary we published with the #preprint!
teapawt.org/lab_COSMO/status/17202โฆ
Tackling excited-state #machinelearning predictions with a hybrid model: @ECignoni and @Divya_Suman14 combined ML predictions of molecular Hamiltonians with time-dependent linear response!
doi.org/10.1021/acscentsci.3โฆ
Fantastic collaboration with @lab_COSMO!
Joseph W. Abbott retweeted
Excited to have kicked off our new course - Practical Programming in Chemistry - for 2nd-year BA chemistry students! #compchem
Huge thanks to our TAs Rebecca (@RebeccaNeeser), Sarina (@SarinaKopf), Daniel (@d_armstr), and Joe (@jwasci) for their help in getting it launched!
Joseph W. Abbott retweeted
Scientists in @MicheleCeriotti's lab at @EPFL_en have used #MachineLearning to paint a more precise picture of how charge transport happens inย lithium thiophosphate,ย a promising material for solid-state batteries. Here's what they found ๐ nccr-marvel.ch/highlights/maโฆ
A fantastic week at #MLIP23 in Helsinki
Great to spend time with other early-career researchers, learn about methods, applications, software, + discuss the future of ML for materials. Receiving a poster prize was a nice bonus.
Thanks to all the organisers for their hard work!
Congratulations to #MLIP23 poster and talk prize winners, @jwasci Patricia Kรถnig @fhi_mpg_de, Sonia Salomoni @Sorbonne_Univ_ , Marthe Guren @UniOslo, Nicklas รsterbacka @chalmersuniv and Raffaele Cheula @AarhusUni_int
And with that, itโs a wrap for now!
Joseph W. Abbott retweeted
After a painful proofing stage ๐คฌ the latest and greatest from ๐งโ๐ @SanggyuChong , @bm_chiheb and @FedericoGrass, and @VLDGroupOx Joe Morrow & @vl_deringer is out on @JCIM_JCTC . So let us introduce the Local Prediction Rigidity and why you should care.๐งต pubs.acs.org/doi/10.1021/acsโฆ
Joseph W. Abbott retweeted
๐จ Sharing a new #compchem #preprint that is the result of a couple years of hard work by @Divya_Suman14, @ECignoni & @jigyasa_nigam . A milestone in our thinking of how to build hybrid models that combine #MachineLearning and physics-based steps in #QChem arxiv.org/abs/2311.00844
Joseph W. Abbott retweeted
RTs please ! ๐ข I have a fully funded #phdposition (UK students) at @ucl @tyc_london โก๏ธ
ucl.ac.uk/condensed-matter-mโฆ
on phase behaviours of nanoscale materials. Interested in #Quantum, statistical mechanics, #ML and excited to team up with my collaborators @ChemCambridge ?
Excitement ... ๐ฅ... I am soon joining the Physics Department @ucl @uclmaps as a Lecturer (Assistant Professor) !! #newpi of the First Principles Statmech (FPS) group to work at the interface of quantum mechanics, statmech and machine learning. DM/email for #PhDposition
Joseph W. Abbott retweeted
Very excited to invite you to #MLIP23 workshop in wintry magical Finland @AaltoUniversity, 6-11 November! We have an awesome lineup of speakers and a mix of talks+hands-on sessions from ML techniques to applications beyond the PES. mlip-workshop.xyz/home
Apply by 22 September!
Joseph W. Abbott retweeted
Our work on the design of MOFs with tunable electrical conductivity (w/ @sven_rogge, Prof. Jihan Kim) has been featured in a virtual issue of @ChemMater, along with many other great works on the chemistry of MOFs. โบ๏ธ
VI: rb.gy/eu1yo
paper: rb.gy/nk4lg
Joseph W. Abbott retweeted
๐ข PhD position in my group @DTUtweet ๐ข. This @DFF_raad project is concerned with the use of decomposed mean-field theory for accelerating QM/MM simulations of light-matter interactions. Please get in touch for further details. RTs appreciated.
efzu.fa.em2.oraclecloud.com/โฆ
Joseph W. Abbott retweeted
#compchem ists love to define and interpret atoms-in-molecules energy decompositions. #MachineLearning models do that for you, but can you trust the values to mean something? See the latest ๐งโ๐ #preprint, in collab with @VLDGroupOx to find out arxiv.org/abs/2306.15638 ๐งต
Joseph W. Abbott retweeted
New PostDoc position! Join the vibrant LIAC team (@SchwallerGroup) for a project on โML for Direct Air Capture Materialsโ. Develop models to accelerate the discovery of new CO2 adsorbents together with excellent experimental collaborators (@lfim_epfl)!
RT appreciated. #ML4Sus
Our systematic DFT study of noble-gas defects in nanodiamonds โคต๏ธ
Nice to see this work out in JPhysChem - a project I contributed to as an undergrad on exchange in the Li Group @UWChemistry, right at the start of my #compchem journey!
Electronic Structures and Spectroscopic Signatures of Noble Gas Doped #Nanodiamonds
By Xiaosong Li, Peter Pauzauskie, Alessio Petrone et al @UW @UninaIT
๐ Open access in ACS Physical Chemistry Au ๐ go.acs.org/3R3
5
176


























