Edgar Sanchez (@SanchezEdgarM) header
Edgar Sanchez (@SanchezEdgarM) profile picture

πŸ‡©πŸ‡ͺ PhD candidate @MPI_Magdeburg | Alumni @ImperialCollege πŸ‡¬πŸ‡§ @quimica_unam πŸ‡²πŸ‡½| @riiaa_org | ML and GNNs for chemical engineering

Magdeburg, Germany
Joined September 2019
Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)Photo by Edgar Sanchez (@SanchezEdgarM)
Students at Stanford have built alphaXiv, an open discussion forum for arXiv papers. Think of it as 𝕏 for research.
87
1,247
119
9,057
647,387
Looking forward to this reading of the one and only @lawrennd! Thanks go to Christoph too! πŸ™ŒπŸΌπŸ¦Ύ
1
6
1,265
arXiv -> alphaXiv Students at Stanford have built alphaXiv, an open discussion forum for arXiv papers. @askalphaxiv You can post questions and comments directly on top of any arXiv paper by changing arXiv to alphaXiv in any URL!
122
1,801
364
6,861
914,433
If you want to extract data using LLMs, you should check out our review πŸ“ arXiv arxiv.org/abs/2407.16867 πŸ’» hands-on online book matextract.pub/
1/8 πŸ“Š Materials science has a wealth of untapped data in papers, but lacks curated datasets for #ML and innovative materials design.
1
16
1
83
12,387
Edgar Sanchez retweeted
Digital Discovery is my favorite scientific journal. πŸ’› (before its inception, I had no favorite.)
Congrats to @digital_rsc on getting its first impact factor: 6.2! It's a heroic effort to get a new journal in a new field so high on its first evaluation. Some of the wildest ideas from my group and others show up in this bohemian journal, so take a look!
2
2
27
2,562
Very happy that this is out! πŸ₯³πŸ₯³ Towards an Artificial Muse for Science, we want to know how interesting AI-generated research ideas are: arxiv.org/abs/2405.17044 100+ Research Group Leaders (!) from 50+ Max Planck Institutes helped us evaluate! Spearheaded by @GuXuemei Fields of research: All topics studied in the Max Planck Society, including biology, chemistry, physics, math, computer science, medicine, social sciences, humanities. Method: We used 50+ million papers to create a knowledge graph. Selected pairs of researchers -> subgraphs. Find possible good connections between them. Use it as a prompt for GPT. My favorite conclusion: We can predict the interest-level from pure knowledge graph data! Thus can generate much more targeted suggestions in the future. Massive THANKS to the more than 100 experienced researchers who spent time on our AI-generated proposals. Thanks to the many scientists who helped us by analysing the outcomes, before we started evaluating the proposals (especially the RGs at @MPI_Light).
6
48
5
188
26,034
New paper out: 'Challenges in data-based reactor modeling: A critical analysis of purely data-driven and hybrid models for a CSTR case study'. Surprisingly, hybrid models didn't always outperform purely data-driven models. Curious? Check it out here: doi.org/10.1016/j.compchemen…
3
8
520
Edgar Sanchez retweeted
this video of the president of guyana tearing this pretentious bbc journalist a new one is probably the most satisfying video i’ve watched in recent memory
216
1,091
260
8,130
972,778
.@SanchezEdgarM & @SundmacherLab explore applicability of GH-GNNs for predicting infinite dilution activity coefficients in polymer solutions. They show promising results compared to established methods marking a pivotal move for industrial thermodynamics go.acs.org/6Zx
5
7
1,820
πŸŽ‰ Check out our new article on using GNNs for predicting activity coefficients of #Polymer solutions. Available code and data on GitHub! doi.org/10.1021/acs.jpca.3c0… πŸ§ͺπŸ” #MachineLearning #GNNs #OpenAccess
2
5
499
Today is the biggest day ever for developers building with @OpenAI We are releasing new models, API’s, open source models, and more. Full details in 🧡
65
290
47
2,873
796,220
Edgar Sanchez retweeted
This is the way to unlock the next trillion high-quality tokens, currently frozen in textbook pixels that are not LLM-ready. Nougat: an open-source OCR model that accurately scans books with heavy math/scientific notations. It's ages ahead of other open OCR options. Meta is doing extraordinary open-source AI, sometimes without as much fanfare as Llama. My first serious AI research project (back @Columbia, 2012) was to convert chemical engineering PDFs into NLP-ready corpus. I still remember the immense pain of Tesseract, a much older OCR system (github.com/tesseract-ocr/tes…). Now Nougat runs a powerful Swin Transformer backbone and blows the benchmarks out of the water. We're talking about double-digit improvements across all metrics. Now, textbooks are all we need for the next GPT! Website: facebookresearch.github.io/n… Open-source code: github.com/facebookresearch/… Paper "Nougat: Neural Optical Understanding for Academic Documents": arxiv.org/abs/2308.13418
116
727
94
3,837
1,089,286
Edgar Sanchez retweeted
Mark your calendars! RIIAA is back in Quito Feb 19-25, 2024 with AI courses, keynotes & a hackathon! Thrilled to have @sirbayes, @sarahookr, @RubenEVillegas, and Felipe Ovideo join, and more to be announced soon! Register & apply for travel grants: forms.gle/24WKAvfeDQ6SU3GF7
13
1
33
12,937
School is starting up again! Time to brush up or dive deeper on #python? Check out pointbreezepubs.gumroad.com/. There is something for everyone.
1
2
3
329
Edgar Sanchez retweeted
Today, we are excited to introduce Graphium into the datamol.io ecosystem! Graphium is a powerful and flexible Python library for training molecular GNNs at scale.πŸš€β™ΎοΈ Try it out today: github.com/datamol-io/graphi…
1
13
2
39
11,212
For everybody who missed it, here is the GitHub repo with codes and practices: github.com/Leticia-maria/Int… 😊 I will be continuously improving and updating it for next versions of the course #compchem ❀️ RTs appreciated and ⭐️ on GH are very welcome.
1
7
36
2,460
Edgar Sanchez retweeted
πŸŽ‰ Join us for our 1st virtual LatinXinChE symposium! πŸ—“οΈ Save the date: Wednesday, Sept/6/2023, 8amPT/11am ET. ⏰ Register here now to attend and be part of our vibrant virtual community bit.ly/3qZl1Ry 🌍 Let's connect, engage, and empower each other in the world of ChE.
26
6
34
15,903