Rhys McAlister

@0xexpt do you need a different pen for this or does it work with regular pens?

This is a long awaited addition I wanted for peptaura, because I believe this is a much superior format to vials! PENS / CARTRIDGES !! The supplier we onboarded sells 3 options : - single-chamber powdered cartridges - single-chamber reconstituted cartridges - dual-chamber cartridges (the best) A dual-chamber cartridge is essentially a cartridge with one chamber with the still lyophilized peptide and another chamber with BAC water. It only mixes/reconstitutes when you insert into the pen, which makes it perfect for shipping! Really excited for this, expect it to be available tonight on EST!

@punishedfounder @ShelB00000 @_TrueVoodoo Can’t dm don’t have premium, is it safe to post here or any alternative way to message?

@aqollo_ @ShelB00000 @_TrueVoodoo it sometimes takes a few days, but it can be really fast we just implemented a feature that helps them process orders better so this should get faster if you DM me your order # I can check for you

Don’t get raped by fitness influencers.

@ShelB00000 @punishedfounder @_TrueVoodoo Don’t have premium can’t dm, is it safe to post here?

@ShelB00000 @punishedfounder @_TrueVoodoo How long after ordering before you received shipping information? Ordered a few days ago and haven’t heard anything yet

@punishedfounder Recommended retailers on the marketplace?

if peptaura becomes the go to marketplace for petides, our leverage will increase to the point where we’ll be able to strong-arm sellers into more stringent QC standards and testing we’ve already caused global peptide prices to drop now we want even better quality

@DrNadolsky Link to the data?

Everyone today when the first retatrutide phase 3 data came out

@HamelHusain @jeremyphoward @juliusai Is there any public info on what’s happening with solveit?

Who is doing interesting integrations between AI and notebooks (like the Jupyter kind, for coding)? Looking for deep integrations that are beyond naive chat with your notebook. Here’s what I know of so far @jeremyphoward : solveit @juliusai : data analysis Who else??

@jeremyphoward Unfortunate to hear, looked like an interesting course

@aqollo_ nup

Management: You know what the world really needs? Devs: ? Management: A new vscode fork. Devs: ??? Management: OK get to work.

@eigensteve @CoryMSimon Interested in helping out

New Book & Video Series!!! (late 2025) Optimization Bootcamp: Applications in Machine Learning, Control, and Inverse Problems Comment for a sneak peak to help proofread and I'll DM (proof reading, typos, HW problems, all get acknowledgment in book!)

@art2san @Splitgate deadlock clips when?

octo-oblivion @Splitgate

@athenamakes Looks great. Is it available to the public?

Built a plugin that turns Obsidian notes into flashcards for spaced repetition📝 Revisiting 3 years of book notes feels like catching up with old friends - familiar, surprising, and full of things I'd forgotten I loved ☺️

@AriWagen @owl_posting @CorinWagen Sounds great, looking forward to it.

@aqollo_ @owl_posting @CorinWagen This is in progress! I've been writing ~1 section/day; it shouldn't take more than a week or so to wrap up.

Can AI improve the current state of molecular simulation? owlposting.com/p/can-ai-impro… in my first podcast, I spend 2 hours interviewing @CorinWagen and @AriWagen, two brothers who are building the next generation of molecular simulation for drug discovery and material sciences 🧵

@AriWagen @owl_posting @CorinWagen That would be fantastic, thanks in advance 🙏

@aqollo_ @owl_posting @CorinWagen Not that I know of, but it’s a great idea! @CorinWagen and I can take a crack at this tomorrow …

@OrbMaterials @yilunliao @MIT @ask1729 @UCBerkeley Will these seminars be made available online for those in other time zones?

Join us this Friday as Yi-Lun Liao (@yilunliao) from @MIT & Aditi Krishnapriyan (@ask1729) from @UCBerkeley explore the new methodologies for training and scaling neural network potentials (NNPs) that are reshaping #AI applications in materials science. bit.ly/3NWwFVN

@CorinWagen @olexandr Hi Corin, slightly off topic but I was wondering if you kept track of timings for calculations internally? It would be interesting to look over a dataset of calculation timings with methods and basis sets and molecule type data.

@olexandr I think the plan is to add more test sets—any suggestions? Agreed that some of the subsets of GMTKN55 are a bit odd, but tough to find CCSD(T)/CBS data and it's nice to be able to compare pretrained NNPs to DFT methods directly. We're working on new benchmarks internally...

While it’s excellent idea to have a leaderboard for benchmarking MLIPS. IMO GMTKN55 is not a good set for that. Starting from the biases in the systems sizes, types to very confusing units;)) 🤓

@HannesStaerk @n_gao96 @n_gao96 Thanks for the excellent presentation, I was wondering if you could share the code you used to prepare the heatmap plot for your paper? Thanks 🙏

Monday with @n_gao96 in the reading group "Learning Equivariant Non-Local Electron Density Functionals" arxiv.org/abs/2410.07972 Join us on zoom at 9am PT / 12pm ET / 6pm CET: portal.valencelabs.com/logg

@gil2rok @remilouf Do you have the link to this post? ty in advance if so

"we don't know what to do with those chains yet, but I have the gut feeling that once we do, probabilistic programming will have another breakthrough" - @remilouf on parallelizing 100s of MCMC chains with JAX

@cloneofsimo How do you usually figure out the optimal hparams other than just grid search?

Wow shampoo epsilon of 1e-1 is very shit: almost as bad as adamW meanwhile, shampoo of 1e-8 is really good. Just realizing again Shampoo epsilon is really really important hparam.