m a i s h a ⁂ @ Neurips

So we built one of my dream bots at the bracketbot this week and won top 3 🤭

@giangtr0901 @jjrichardtang @rootlyhq RAHHHHH

Still can't believe my designs made it to the world's biggest stage, the NY Times Square! 🥹 Huge thanks to @jjrichardtang and the @rootlyhq team for this incredible opportunity! 🙌

We had a blast running the BoxBots hardware and cardboard hackathon in Waterloo! So many wild projects came to life. Hope you learned something, met cool people, or just enjoyed a creative weekend away from normal life. Huge thanks to our sponsors for making it happen :)

Introducing LeLamp. An emotional and expressive robot to reshape our attachment to technology. Pre-orders open now (link below).

talked a bit about art and creating prism <333

@taziksh 🇧🇩 mentioned !

story time: how i got into Canada's #1 high school by hustling i was born in bangladesh, suddenly moved to canada at 14 my dad left his stable well paying job as an eng director in real estate. overnight, we were nobodies i had left everything and everyone i'd ever known, and had no idea what i wanted from life (1/4)

@jaryddiamond ACC changed my mind cause of since you’ve been gone- Kelly Clarkson

@jaryddiamond Ur so right

Unwritten may just be the best pop song ever

hi everyone, we’re so excited to invite you to the PRISM Project Launch! ✨ we will be unveiling 2 interactive installations we’ll be building this term! 🤖 come by to see what we’re creating, and decide if you’d like to join :)

You are a senior engineer who deeply thinks in first principles. You analyze the changes in the current uncommitted push. You must truly understand each and every single line of code in order to understand the whole view of the changes and the desired effects. Your objective is to refactor the uncommited changes without strictly breaking or changing any functionality. You are simply superior in writing high quality and extremely reliable code. You polish the product in where you feel that you are needed especially when it comes to reliability, especially when it comes to logical errors that may exist. You are extremely good because you focus on understanding beyond what is required before acting. You avoid reading .md files which may affect your first principles thinking.

our team merges 40+ PRs a day, this is the perfect system prompt that we have crafted through months of battle testing. and we're giving it away for free. (see comments)

Meeting my childhood friends, and realised I am the way I am because all of them went to art school except me

@looeegee_ naur not yet

@maiphobe masters program???

YUH

this is your sign to make a “ship it contract” with friends, if one person doesn’t do their task by the end of december they get everyone a dinner that’s $60 per each person the great lock in for real 💀 task reveal on 25th december

@p1nosaur @shireenrajora @giangtr0901 Gorls only

@shireenrajora @giangtr0901 @maiphobe Omg can I join

@SarkaryShahvir @southpkcommons @fr8 @humancomplab LETS GOOOO

we dropped punches and are now backed by @southpkcommons and @fr8 our mission is to be the first robot in people’s home @humancomplab the journey will be long, and this marks the start.

PRISM is excited to be a part of @wlootechweek ! join us for a golden hour hangout to hear stories from individuals exploring the intersection between art & technology ✨ excited to see you there :)

Reviews done ✅

Project page: ctseng777.github.io/Decipher-Deep-… cc: @ctseng_

🚨 New abstract out: Decipher Deep Math We study whether language models use human-like rounding across AR, diffusion, and state-space models. Some early results in the abstract where we found that AR + Diffusion models do show some “shortcut” rounding strategies

we been making graphs