Sign in

Arnav Das

@arnav7.bsky.social
2K followers 13K following 1.2K posts

My posts here will be a mishmash of personal opinions and smol curious excursions..

PostsRepliesMedia
Reposted by Arnav Das
Ann-Sophie Barwich @smellosopher.bsky.social · 14h
"In 2004...Zhuo-Hua Pan achieved something remarkable: he made neurons react to light...a foundational step in creating optogenetics—that would go on to revolutionize neuroscience. ...you’ve likely never heard his name. Instead, the world knows Karl Deisseroth and Ed Boyden." lnkd.in/p/ggR8fvY7
lnkd.in
He May Have Invented One of Neuroscience's Biggest Advances--but You've Never Heard of Him | Mikhail Lebedev
The "Invisible" Pioneer: How one scientist missed the Nobel-winning spotlight by 6 months. In 2004, neuroscientist Zhuo-Hua Pan achieved something remarkable: he made neurons react to light. This was...
061
Reposted by Arnav Das
Ben Brubaker @benbenbrubaker.bsky.social · 11h
OpenAI’s big math dump last night included a proof of the unique games conjecture, one of the most famous open question in complexity theory. Three weeks ago, three computer scientists settled a closely related conjecture “the old-fashioned way.” My latest for @quantamagazine.org:
quantamagazine.org
As AI Closed In on ‘Unique Games’ Proof, Researchers Raced to Beat the Machines | Quanta Magazine
In the shadow of a rumored AI proof of one of the biggest problems in their field, three computer scientists rushed to publish their own milestone result.
0135
Reposted by Arnav Das
koenfucius @koenfucius.bsky.social · 15h
Does ‘doing something together’ require that the participants adhere to norms governing the joint action? Experimental research by @jglavin.bsky.social & Rachar suggests the answer is yes—joint action does imply normative relations: buff.ly/zUbqGN0 HT @xphilosopher.bsky.social
031
Reposted by Arnav Das
Sam Power @spmontecarlo.bsky.social · 12h
The basic idea: write down the training dynamics of a simple NN, summarise them with a two-dimensional system of differential inequalities, solve these inequalities, and deduce things as you would like. All pretty elementary, more-or-less teachable, somewhat extensible to other non-convex problems.
111
Reposted by Arnav Das
Sam Power @spmontecarlo.bsky.social · 12h
I finally got around to completing a note that's been on the side for a month or two now: "Optimisation of Wide, Shallow Neural Networks by Gradient Flow" github.com/sampower88/P...
1101
Reposted by Arnav Das
Maria Antoniak @mariaa.bsky.social · 10h
It's infuriating how the U.S. is treating our international researchers. We are so incredibly lucky that talented people want to come to our country to study, research, work, and build. Trump, Musk, and their cronies are destroying the best parts of science in the U.S.
410217
Reposted by Arnav Das
Julia M. Rohrer @dingdingpeng.the100.ci · 9h
Still upset about that highly cited paper in Psych Science you consider deeply flawed? Now is your time to shine ✨
66928
Reposted by Arnav Das
koenfucius @koenfucius.bsky.social · 9h
“Hundreds of clinical trials now show that exercise can significantly reduce symptoms of depression.” Shouldn’t it be acknowledged and treated like a medicine? Yes, argues Nicholas Fabiano: buff.ly/A0R9u74
034
Reposted by Arnav Das
Vincent Carchidi @vcarchidi.bsky.social · 07/10/2026
OpenAI released 700 math solutions today. I went through them so you don't have to. Here's what you need to know 👇
4441
Reposted by Arnav Das
Kenny Peng @kennypeng.bsky.social · 06/10/2026
Our new paper introduces a scientific theory of atomic features. We mathematically derive testable predictions. Our experiments challenge conventional wisdom. SAEs of different size and training data share many features. Large SAEs recover both parent and child features. 🧵
1207
Reposted by Arnav Das
Terence Tao @teorth.bsky.social · 06/10/2026
I came across this parody of a rumored upcoming press release by OpenAI: "OpenAI Releases the Final Ten Minutes of 500 Previously Unreleased Films, Ushering in a New Era of Movie Watching" : x.com/Endings500/s...
x.com
Endings Fivehundred (@Endings500) on X
FOR IMMEDIATE RELEASE: October 6th, 2026 OpenAI Releases the Final Ten Minutes of 500 Previously Unreleased Films, Ushering in a New Era of Movie Watching
518237
Reposted by Arnav Das
Grace @gracekind.net · 07/10/2026
Holy shit
Openai github repository named "critical-theory"
2354664
Reposted by Arnav Das
Benjamin @benjaminschneider.ch · 06/10/2026
„medically“
091
Reposted by Arnav Das
Tim Verstynen @tdverstynen.bsky.social · 06/10/2026
As my neighbors are trying to figure out if they were exposed to the measles while shopping this weekend, I’m pondering where to place Andrew Wakefield in the pantheon of mass murderers. www.wtae.com/article/penn...
wtae.com
Measles case confirmed in Allegheny County with possible exposure at Target, health dept. says
An Allegheny County resident has tested positive for measles, with potential exposure reported at a Target store, health officials said.
173
Reposted by Arnav Das
Vaishnavi S @vaishs.bsky.social · 07/10/2026
Most iconic footnote in the history of footnotes
081
Reposted by Arnav Das
Michael Hobbes @michaelhobbes.bsky.social · 06/10/2026
According to a 2013 paper, almost 40% of female homicides globally are due to intimate partner violence www.thelancet.com/journals/lan...
12591100
Reposted by Arnav Das
Freya Holmér @freya.bsky.social · 06/10/2026
the "adapt or die" people failed to consider the secret third option of forming a neoluddite community where I'm selling my latest video game on usb sticks in a little lemonade stand
472273306
Reposted by Arnav Das
Cat Hicks @grimalkina.bsky.social · 06/10/2026
I am reading about organizational justice, fairness and unfairness, scholarship I've been reading for years, for the next Fight for the Human newsletter. I keep trying to make these shorter but I am who I am! This is a fascinating literature I can't wait to spotlight it. www.fightforthehuman.com
fightforthehuman.com
Fight for the Human by Cat Hicks
Rehumanizing tech with a psychologist for the humans of tech
2121
Reposted by Arnav Das
Zeta Of 1 @zetaof1.bsky.social · 06/10/2026
One thing that I find very intriguing is that LLMs have, at least to me, resurfaced an interesting point by Immanuel Kant that I ten years ago believed to be ridiculous but now feel has an application. To Kant, only humans have capacity for moral action and are also appropriate targets for moral 1/
131
Reposted by Arnav Das
David Marx @digthatdata.bsky.social · 06/10/2026
An important component of how republicans maintain narrative control is through Orwellian choices for naming their ideologies. "originalism" is historical revisionism. "pro-life" asserts that women's lives have less value than their unborn children. "conservatism" dismantles the status quo.
0206
Reposted by Arnav Das
Ann-Sophie Barwich @smellosopher.bsky.social · 06/10/2026
Piece by @edyong209.bsky.social on the Nobel in 2017. May as well be 2026, though. "They distort the nature of the scientific enterprise, rewrite its history, and overlook many of its most important contributors." www.theatlantic.com/science/arch...
theatlantic.com
The Absurdity of the Nobel Prizes in Science
They distort the nature of the scientific enterprise, rewrite its history, and overlook many of its most important contributors.
061
Reposted by Arnav Das
Berna Devezer @devezer.bsky.social · 05/10/2026
what i understand from this thread is LLMs, just like most of the literature, are confused about what constitutes p-hacking, what is cherry picking, and what is HARKing. you can actually do any of these things independently from the others. i'd also love to see the first claim in action.
1293
Reposted by Arnav Das
Ann-Sophie Barwich @smellosopher.bsky.social · 06/10/2026
*Any* movie with professors sitting in comfy office chairs near windows, musing about the world. No, you write emails. A window means you got tenure. You cannot open said windows. Because your administration fears you may jump. Not that they care. But the paperwork resulting from this... oh gawd.
2405
Reposted by Arnav Das
Janel Le Belle @janellebelle.bsky.social · 06/10/2026
Human brain assembloids as a model of anaesthetic-induced neural dynamics in vitro www.sciencedirect.com/science/arti...
sciencedirect.com
011
Reposted by Arnav Das
James @nonzerosum.games · 06/10/2026
These posts are a contracted "Dev Log" showing the many iterations over the last 10 years for my upcoming game NEUTROPIA which teaches non-zero dynamics. The 1st version was based on light-switches... they made little clicky noises. #GameDev #Games #MobileGames #GamesForChange
0115
Reposted by Arnav Das
Olivia Guest · Ολίβια Γκεστ @olivia.science · 06/10/2026
USAmericans insisting they are the victims while starting wars and doing genocides for a century plus while also being the richest country. It's so disturbing to me we're all so confused about the relationships between governments and their people. Normalised apathy.
38517
Reposted by Arnav Das
Ann-Sophie Barwich @smellosopher.bsky.social · 06/10/2026
Watson is the Sam Altman of Genetics.
1197
Reposted by Arnav Das
koenfucius @koenfucius.bsky.social · 06/10/2026
We age, and our brain with us, but there are several things we can do (or not do!) to keep our cognitive abilities up to scratch as we get older: buff.ly/ug420yJ
011
Reposted by Arnav Das
Darren Dahly @statsepi.bsky.social · 06/10/2026
They hounded Aaron Swartz to death for downloading papers from jstor but society is now supposed to just accept mass theft because this thing can write that TPS report *for* you. en.wikipedia.org/wiki/Aaron_S...
en.wikipedia.org
Aaron Swartz - Wikipedia
111558528
Reposted by Arnav Das
koenfucius @koenfucius.bsky.social · 05/10/2026
Our biggest strengths can become our biggest problems—they tend to entrench the way we analyse and solve problems. The key to break out of this cognitive entrenchment is curiosity, writes @neuroscienceof.bsky.social: buff.ly/rDZx4JJ
011
Reposted by Arnav Das
Clément Canonne @ccanonne.github.io · 05/10/2026
Moni Naor is awarded the Knuth Prize, for his incredible body of work, spanning cryptography, pseudorandomness, algorithms, and much more! tc.computer.org/tcmf/2026/10... h/t @gautamkamath.com
tc.computer.org
2026 Knuth Prize is awarded to Moni Naor - IEEE Computer Society Technical Committee on Mathematical Foundations of Computing
Moni Naor is the recipient of the 2026 Donald E. Knuth Prize.The Knuth prize honors sustained and profound contributions to thefoundations of computer science, and it is difficult toimagine a career t...
0308
Reposted by Arnav Das
Dr Mircea Zloteanu 🌺🌞🍃 @mzloteanu.bsky.social · 05/10/2026
#statstab #631 Prior Information in Frequentist Research Designs: The Case of Neyman’s Sampling Theory Thoughts: Model-based v design-based problem. This should be discussed more. #modelbased #designbased #sampling #priors #neyman #frequentist #bayesian link.springer.com/article/10.1...
link.springer.com
Prior Information in Frequentist Research Designs: The Case of Neyman’s Sampling Theory - Journal for General Philosophy of Science
We analyse the issue of using prior information in frequentist statistical inference. For that purpose, we scrutinise different kinds of sampling designs in Jerzy Neyman’s theory to reveal a variety o...
031
Reposted by Arnav Das
Mark Riedl @markriedl.bsky.social · 06/10/2026
The Capabilibara project seeks to understand how a language model learns to interpret people's beliefs, emotions, intentions, and everyday moral choices. We trace that ability back to the training data using influence functions and unlearning hcai-lab-gt.github.io/capabilibara/ Find us at COLM!
1318
Reposted by Arnav Das
koenfucius @koenfucius.bsky.social · 06/10/2026
Research by Parton et al suggests sadism (actively inflicting pain on others for pleasure) is correlated, not just with schadenfreude (passively enjoying others’ misery), but also with greater *habituation* to schadenfreude (=declining pleasure): buff.ly/ohK833F
011
Reposted by Arnav Das
James @nonzerosum.games · 06/10/2026
Oh, and this is the first version... from 2014. Look at those auto-shadows... amazing. I actually did like making the dungeon tiles, they're all hand drawn (9 of them repeated and rotated) I'll surely use them for something one day. #GameDev #DevLog
0229
Reposted by Arnav Das
Lucy D’Agostino McGowan @lucystats.bsky.social · 05/10/2026
🎙️ Today on Regression to the Mean Girls @sarahlotspeich.bsky.social & I chat with @mattansb.msbstats.info about a marshmallow test (no, not that one!) This ep has everything! Bayesian models! Food science! A twist at the end! Listen wherever you listen to podcasts!
2297
Reposted by Arnav Das
Christian Wolf @chriswolfvision.bsky.social · 06/10/2026
Having fun with Blender ... Distilling context transformers into recurrent transformers: arxiv.org/abs/2606.21562 @weinzaepfelp.bsky.social @mbsariyildiz.bsky.social
1225
Reposted by Arnav Das
koenfucius @koenfucius.bsky.social · 06/10/2026
Planning ahead can be seen as a fair advantage, but knowing everyone else’s job, that’s an unfair advantage—and Jim Clark, 61 years ago, had both, writes Dave Trott: buff.ly/z14lg6d
021
Reposted by Arnav Das
Sean Carroll @seanmcarroll.bsky.social · 06/10/2026
Such a good Nobel Prize! Francis Halzen is a great person, and “let’s turn the Antarctic ice sheet into a neutrino detector” is the kind of idea that deserves a Prize if you can pull it off.
217829
Reposted by Arnav Das
Prof. Dr Beth Singler @bvlsingler.bsky.social · 06/10/2026
Resident Evil...
static.klipy.com
Chef'S Kiss Amanda Seyfried
ALT: Chef'S Kiss Amanda Seyfried
021
Reposted by Arnav Das
Vera Wilde @verawil.de · 05/10/2026
Causal discovery. What is it good for? I've criticized the Glymour–Spirtes Tetrad tradition before; NOTEARS is its younger cousin. I ran a simulation showing if you rescale one variable, the "causes" change. Causality first! Analyses second. wildetruth.substack.com/p/no-tears-f...
2195
Reposted by Arnav Das
Justin Weinberg @dailynous.com · 06/10/2026
If you could give beginning philosophy graduate students just one piece of advice, what would it be? (If you're tempted to respond "don't", don't.)
dailynous.com
Your One Piece of Philosophy Grad School Advice - Daily Nous
If you give beginning philosophy graduate students just one piece of advice, what would it be? I know that some readers will want to jump […]
3102
Reposted by Arnav Das
M.J. Crockett @mjcrockett.bsky.social · 06/10/2026
Excellent piece www.insidehighered.com/opinion/view...
insidehighered.com
On Proper Objects of Study and the Humanities (opinion)
A recent report on the humanities criticizes the distortion of scholarship in the service of political goals. Other disciplines would like a word.
032
Reposted by Arnav Das
Sarah Wieten @sarahwieten.bsky.social · 06/10/2026
If anybody reading this is a working scientist who would be interested to learn about and co-develop a public philosophy of science course for scientists, dm me, thanks! Shares appreciated!
12322
Reposted by Arnav Das
Ben Recht @beenwrekt.bsky.social · 06/10/2026
The obsession with and futility of macroeconomic forecasting.
argmin.net
Too Big to Fail
The obsession with and futility of macroeconomic forecasting
182
Reposted by Arnav Das
Cat Hicks @grimalkina.bsky.social · 06/10/2026
because tasks gate other tasks. Because measuring the hours saved by washing machines is a relatively useful index for quality of life improvement. Because the time it takes for you to get from the ER to the operating room is a predictor. A billion different task-based answers to this one.
11888
Reposted by Arnav Das
Cat Hicks @grimalkina.bsky.social · 06/10/2026
"why would I ever need a task to be done faster" come on, guys. Because we're dying. Because we are literally all dying all the time. Because we're running out of time. The fact that it's often hard to measure or stupidly measured doesn't mean "saving time" isn't an important thing to human beings
1647758
Reposted by Arnav Das
koenfucius @koenfucius.bsky.social · 06/10/2026
How much should we care for insects—compared to humans? Most would say less. Should we care more about many insects than about one? Most would agree. So if there are enough of them, we should care more about them. Or might the numbers not add up? buff.ly/DrtTkjA
011
Reposted by Arnav Das
Berna Devezer @devezer.bsky.social · 05/10/2026
came across this fact while doing crossword last night: wombats poop perfect cubes. how come i did not know this before?! here to fix that for you, just in case you didn't either.
9 cubical wombat poops in slightly varying sizes Four cubical wombat poops, almost identical looking, on grass
5243
Reposted by Arnav Das
Stephen Wild @stephenjwild.bsky.social · 05/10/2026
This is a fantastic graph because the increase in "married men and women having an affair" number and the post-2022 drop gives a clue that something may be up (probably something to do with the 2022 numbers). It would be great to have the numbers for additional years.
261