🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

And perhaps most importantly, may you not be a citation involving one of the anonymous reviewers.

plan: 2019: be ready & fully caffeinated the millisecond it goes up 2020: write some JS to register me instantly 2021: discover the physical location of the NIPS backend & plug in via fibre optic next door 2022: orchestrate a DDOS attack of requests for my registration 2023: ???

2023+: only those with access to powerful AGIs can now register for the NIPS conference.

More than 100 projects at the poster session of Stanford University CS224W Mining and Learning with Graphs, organized by Jure Leskovec and yours truly! Check the reports here: web.stanford.edu/class/cs224w/p…

A friendly reminder to all the physicists (and not only) out there. Stop writing that paper. Cut that shit out. #COVID19 #yourmodeliswrong reddit.com/r/Physics/comm… h/t Michele Starnini

Jozsef Konczer: Statistical Games arxiv.org/abs/2402.15892 arxiv.org/pdf/2402.15892

What Alzheimer noted as fat saccules in glia may have been lipid droplets. They can be induced by Abeta in microglia leading them to release factors that damage neurons. ApoE4 makes things worse rdcu.be/dA6Kh. Congrats Michael Haney Robert Palovics and fantastic team

biorxiv.org/content/10.110… Nice benchmark of single cell "foundation models" (scGPT, scFoundation) and GEARS (a GNN model) further hyped as "virtual cell models" against linear baselines on perturbation prediction. Long-story short: they can't beat the linear baselines. 1/

Your language model is wasting half of its layers to just refine probability distributions rather than doing interesting computations. In our paper, we found that the second half of the layers of the Llama 3 models have minimal effect on future computations. 1/6

Rossz úton lépegetett, elfogtuk. Nem ijedünk meg a sötét oldal árnyékától. Star Wars #StarWars8 #TheLastJedi #police #rendőrség

Hello, Ph.D students! I bring news from professor land. You know how when you’re doing research you constantly feel like you have no clue what you’re supposed to be doing? Turns out that doesn’t go away when you become a professor. We just act all confident when we advise you.

LOL, így volt ez tegnap Zalában is, kár hogy a tisztelt kollégák szerint a ködfényszóró használatához az ellenőrzött szakaszon nem volt elég szignifikáns az amúgy fél megyét beterítő köd. x.com/police_HU/stat…

When Matt Slack is better than your missile defence warning system...

For anyone who missed the #NIPS2017 tutorials, all of the videos are now up! nips.cc/Conferences/20…

If you're learning TensorFlow, highly recommend checking out Eager Execution (code: goo.gl/Hp2hWA, slides: goo.gl/74t66Q). This makes it *much* easier to understand how various functions work, because you can execute and debug them interactively.

Having spent Valentine's Day alone, the grad student has a fiery rendezvous with heavily discounted post-holiday candy.

Binary search, merge sort, public key cryptography, and other algorithms explained as IKEA furniture assembly instructions idea-instructions.com by @ideainstruction

Important words from Fan Chung Graham. x.com/SebiCioaba/sta…

Now @aaronclauset.bsky.social being introduced (very nicely) by Albert-László Barabási I feel the room is full of love #complenet18 #complenet 𝙲𝚘𝚖𝚙𝚕𝚎𝙽𝚎𝚝 𝟸𝟶𝟸𝟼

Asking a committee member whether he is ready for the job market, the grad student suffocates in the hesitant silence.