Gen-AI protein with varying loop length, sequence, and structure. In the context of an antibody and antigen complex, this could be used to generate novel binders for a particular epitope.
It's mindboggling that many years of research, engineering, community engagement, and meticulous benchmarking all distill into a single, elegant, plug-and-play function:
torch.compile()
You rarely get to see so much behind the scene in a single blog. https://t.co/JWLqFo0vcI
new paper who dis
@demishassabis talked about "delusions and debugging" during a lecture on self-learning systems and that inspired our submission to the UpML workshop at @icmlconf 2022:
https://t.co/LWyMlUNwsQ
#ICML2022@uwaterlooAI@UWaterloo
What team is doing the best work on using machine learning for intelligent design of materials?
e.g. @Mitra_Chem, Chemix, @QuantumScapeCo, and @SilaNanotech. Would love to know which industry teams/academic labs are aggressively pushing for ml for accelerated material discovery!
What team is doing the best work on using machine learning for intelligent design of materials?
e.g. @Mitra_Chem, Chemix, @QuantumScapeCo, and @SilaNanotech. Would love to know which industry teams/academic labs are aggressively pushing for ml for accelerated material discovery!