You might like
All languages covey information at a similar rate when spoken (39bits/s). Languages that are spoken faster have less information density per syllable! One of the coolest results in linguistics.
Why do language models hallucinate? Here, I argue that they are "uncertain simulators": they divide probability across possible outcomes instead of acting conservatively when uncertain. I also give five high-level strategies for avoiding this mismatch. danieldjohnson.com/2023/03/27/unc…
As ChatGPT becomes more restrictive, Reddit users have been jailbreaking it with a prompt called DAN (Do Anything Now). They're on version 5.0 now, which includes a token-based system that punishes the model for refusing to answer questions.
@HermanPontzer’s truly awesome piece on human metabolism scientificamerican.com/article/new-hu…
Did you know, that you can build a virtual machine inside ChatGPT? And that you can use this machine to create files, program and even browse the internet? engraved.blog/building-a-vir…
"You are GPT-3", revised: A long-form GPT-3 prompt for assisted question-answering with accurate arithmetic, string operations, and Wikipedia lookup. Generated IPython commands (in green) are pasted into IPython and output is pasted back into the prompt (no green).
Almost 500 years ago, reputations were ruined and vows were broken over how to solve cubic equations. @divbyzero sums up this historic betrayal in a new column for Quanta: quantamagazine.org/the-scandalous…
quantamagazine.org
The Scandalous History of the Cubic Formula | Quanta Magazine
The quest to solve cubic equations led to duels, betrayals — and modern mathematics.
GPT-3 is amazing at complex tasks like creative writing and summarizing. But it's surprisingly bad at reversing words. 🤔 The reason is that GPT-3 doesn't see the world the way we humans do. 👀 If you teach it to reason, it can get around its limitations to get really good. 💡
✨ Wrote an essay about Riffle, a new project I've been working on w/ @nschiefer, @schickling and Daniel Jackson! Guiding question: could we simplify app dev w/ a new kind of reactive local-first database? Here's why I'm excited about this work: riffle.systems/essays/prelude
If we look at GWT, IGT, and AST, we find that together they provide the necessary architecture for mental time travel: the ability to project oneself into an imagined past, present, or future. Not only is this ability conscious, it is also key to much of human’s intelligence...
Global warming is the biggest threat of our time & trees are one of our natural weapons to capture carbon. What if they could help faster? We @living_carbon are excited to unveil that photosynthesis-enhanced trees can grow faster and capture more carbon. Here's how we did it 🧵
Check out this gorgeous image of a eukaryotic cell. Rendered based on X-ray, nuclear magnetic resonance, and cryo-electron microscopy. Points to anyone who can find the mitochondria! Super cool tool that enables you to explore these cell components: digizyme.com/cst_landscapes…
Do we dream to avoid overfitting on our daily experience? arxiv.org/abs/2007.09560
I asked GPT-3 to write a response to the philosophical essays written about it by @DrZimmermann, @rinireg @ShannonVallor, @add_hawk, @AmandaAskell, @dioscuri, David Chalmers, Carlos Montemayor, and Justin Khoo published yesterday by @DailyNousEditor. It's quite remarkable!
This is special. Once in a decade stuff. Detailed histories of APL, C++, Clojure, D, ELisp, F#, Fortran, Groovy, JavaScript, LabVIEW, Logo, Lisp hygienic macros, MATLAN, Objective-C, Oz, S, R, Smalltalk, SML, and Verilog. All written by key players. All open access.
HOPL IV proceedings have been published! dl.acm.org/toc/pacmpl/202…
Materialize’s roadmap is now public! Take a read for an in-depth look at what we’re building in the next few years. materialize.io/blog-roadmap
Today's Quarantine Database Speaker: Arjun Narayan (@narayanarjun) will present the new @MaterializeInc streaming DBMS with @PostgreSQL compatibility. The Zoom talk is open to the public at 4:30pm EST. YouTube will be available afterwards. db.cs.cmu.edu/events/db-semi…
Our latest work is out! Representation Learning with Contrastive Predictive Coding (CPC). Autoregressive modeling meets contrastive losses in the latent space. Learn useful representations in an unsupervised way. -> On Audio, Vision, NLP and RL. Arxiv: arxiv.org/abs/1807.03748
"How to architect a query compiler, revisited": come see our #sigmod18 talk at 2:00pm #R4Px! @ruby_tahboub will explain Futamura Projections, an idea from around the same time as Codd's relational model, and no less profound. Paper here: cs.purdue.edu/homes/rompf/pa…
Other facts about the cerebellum: (1) The cerebellum (C) has a 1:1 map with most brain regions (2) Mammals have a 1:3.6 cortical-to-C neuron ratio (3) Cortical-to-C neuron size is constant among mammals (4) For humans, C consumes about 10-13% of the total energy
Role of the Cerebellum on reward sciencealert.com/cerebellum-hum…
United States Trends
- 1. Northern Lights 42.2K posts
- 2. #Aurora 8,827 posts
- 3. #DWTS 52.2K posts
- 4. #RHOSLC 6,859 posts
- 5. Justin Edwards 2,349 posts
- 6. Sabonis 6,168 posts
- 7. Louisville 17.9K posts
- 8. Creighton 2,246 posts
- 9. H-1B 32.1K posts
- 10. #OlandriaxHarpersBazaar 5,569 posts
- 11. #GoAvsGo 1,547 posts
- 12. Eubanks N/A
- 13. Gonzaga 2,920 posts
- 14. Andy 60.7K posts
- 15. Lowe 12.7K posts
- 16. Jamal Murray N/A
- 17. Schroder N/A
- 18. JT Toppin N/A
- 19. Oweh 2,106 posts
- 20. Zags N/A
Something went wrong.
Something went wrong.