aolwyn @aolwyncode

Adam. AI / ML MASc candidate. Top 1% video game player in multiple games. Here to learn and improve. Canada Joined July 2016

Tweets

119
Followers

15
Following

257
Likes

425

Rohan Pandey @khoomeik

2 weeks ago

once a scientific paradigm has been established, improving sota is just a matter of puzzle-solving and RL on LLMs is very good at solving puzzles

Aryaman Arora @aryaman2020

2 weeks ago

once a scientific paradigm has been established, improving sota is just a matter of puzzle-solving and RL on LLMs is very good at solving puzzles

1 0 25 17K 7

Download Image

10 8 156 16K 62

Zhengzhong Tu @_vztu

a month ago

Dear NeurIPS reviewers, please be reminded to delete the GPT prompts next time :)

25 61 1K 89K 141

Download Image

Will McGugan @willmcgugan

2 months ago

I'm writing up a blog post about Markdown streaming, because I think it is an interesting topic. I reckon there will be almost a dozen people who agree with me!

90 110 2K 248K 2K

Download Gif

Since it's summer, and more or less internship and tech interview season, I made all 30 chapters of my Machine Learning Q and AI book freely available for the summer: sebastianraschka.com/books/ml-q-and… Hope it’s helpful! Happy reading, and good luck if you are interviewing!

40 323 2K 150K 2K

Hypixel Server @HypixelNetwork

3 months ago

Hypixel SkyBlock Foraging is now available for public testing on the Hypixel Alpha Network! To thank you guys for being so patient, we are giving away 16,400 SkyBlock Gems (approx. $100 value on the Hypixel Network store 👀) to 5 winners! ⭐️ To enter, you must like and retweet…

156 2K 3K 127K 75

Download Image

Xuandong Zhao @xuandongzhao

3 months ago

🚀 Excited to share the most inspiring work I’ve been part of this year: "Learning to Reason without External Rewards" TL;DR: We show that LLMs can learn complex reasoning without access to ground-truth answers, simply by optimizing their own internal sense of confidence. 1/n

84 510 3K 543K 4K

Download Image

neural nets. @cneuralnetwork

3 months ago

today my roommate, who was working on research in another field, handed in his resignation and quit research overall he iterated that research is most probably a 24/7 job, where people just work because they love to do that work all over the day, and the main incentive is always…

11 19 597 39K 111

Anil Seth @anilkseth

4 months ago

1/3 @geoffreyhinton once said that the future depends on some graduate student being suspicious of everything he says (via @lexfridman). He also said was that it was impossible to find biologically plausible approaches to backprop that scale well: radical.vc/geoffrey-hinto….

17 107 976 97K 518

Download Image

Yoni Smolyar @YoniSmolyar

4 months ago

By far the greatest source of anxiety in my life is the voice in my head saying I should be working right now. I should be building/coding/marketing/making videos/researching/experimenting/hunched over my laptop working Ever since I started this solopreneur thing last year…

255 182 3K 464K 1K

vittorio @IterIntellectus

6 months ago

holy shit MIT researchers just turned skin cells directly into neurons without stem cell intermediate, 100-fold efficiency boost, and they actually worked when transplanted into mouse brains 1/

425 2K 17K 2.5M 6K

Download Image

Dexerto @Dexerto

6 months ago

The Pirate Bay co-founder Carl Lundstrom has died in a plane crash at age 64

750 2K 37K 1.7M 3K

Download Image

Theo - t3.gg @theo

6 months ago

This is one of the wildest git diffs I've ever seen

227 2K 17K 909K 2K

Download Image

vittorio @IterIntellectus

7 months ago

HOLY SHIT IT'S HAPPENING AI can now write genomes from scratch. Arc Institute an NVIDIA just published Evo-2, the largest AI model for biology, trained on 9.3 trillion DNA base pairs spanning the entire tree of life. it doesn’t just analyze genomes. it creates them 1/

937 4K 24K 3.5M 14K

Download Image

David @dzhng

7 months ago

Introducing deep-research - my own open source implementation of OpenAI's new Deep Research agent. Get the same capability without paying $200. You can even tweak the behavior of the agent with adjustable breadth and depth. Run it for 5 min or 5 hours, it'll auto adjust.

145 460 5K 518K 6K

Download Video

Kevin Pu @kevpjk

7 months ago

🔬Research ideation is hard: After the spark of a brilliant initial idea, much work is still needed to further develop it into a well-thoughtout project by iteratively expanding and refining the initial idea and grounding it to relevant literature. How can we better support this?

11 82 643 92K 741

Download Image

ludwig @ludwigABAP

8 months ago

either you accept you're not meant for more and you settle at the level you belong at (many such cases) or you reject that notion and do what is necessary to become more there is literally nothing else to it

51 667 6K 332K 2K

Laura Ruis @LauraRuis

10 months ago

How do LLMs learn to reason from data? Are they ~retrieving the answers from parametric knowledge🦜? In our new preprint, we look at the pretraining data and find evidence against this: Procedural knowledge in pretraining drives LLM reasoning ⚙️🔢 🧵⬇️

24 207 965 189K 986

Download Image

sankalp @dejavucoder

10 months ago

i understand you guys like lists LeNet (1989) LSTM (1997) Deep Belief Networks (2006) AlexNet (2012) Word2Vec (2013) GAN (2014) VGG16/19 (2014) InceptionNet (2014) VAE (2014) ResNet (2015) U-Net (2015) YOLO v1 (2015) FastRCNN (2015) FasterRCNN (2015) Neural Style Transfer…

25 122 1K 83K 988

Ishaan @auto_grad_

10 months ago

these are some computer vision papers that everyone must go through atleast once: 1. ResNets: arxiv.org/pdf/1512.03385… 2. YOLO: arxiv.org/abs/1506.02640 3. DeConv: lxu.me/mypapers/dcnn_… 4. GAN: arxiv.org/abs/1406.2661 5. Unet: arxiv.org/abs/1505.04597 6. Focal Loss:…

23 167 1K 69K 2K

Hieu Pham @hyhieu226

10 months ago

Grok-3 just proved Riemann's hypothesis. We decided to pause its training to check its proof, and if the proof is correct, training won't be resumed, as the AI is deemed so smart that it becomes a danger to humanity.