nathan-barry

5 ms·

1.
Can you tell which AI-generated text is watermarked?(watermark-quiz.krishmatta.net)
3 points by nathan-barry 29d ago|0 comments
2.
by nathan-barry 3mo ago
LLMs are very good at lossless compression via arithmetic coding. But I didn't know that it was possible to go the reverse direction (do language modeling via a compressor). It's not super great quality, but I'm surprised it
3.
Can gzip be a language model?(nathan.rs)
15 points by nathan-barry 3mo ago|2 comments
4.
Large Text Compression Benchmark(mattmahoney.net)
3 points by nathan-barry 3mo ago|1 comments
5.
Language Modeling Without Neural Networks(nathan.rs)
2 points by nathan-barry 5mo ago|0 comments
6.
Quines in Every Programming Language (Rosetta Code)(rosettacode.org)
2 points by nathan-barry 5mo ago|0 comments
7.
Mercury Edit 2: Fastest next-edit prediction with a diffusion LLM (221ms)(inceptionlabs.ai)
1 points by nathan-barry 6mo ago|0 comments
8.
Noise schedules considered harmful (2024)(sander.ai)
1 points by nathan-barry 7mo ago|0 comments
9.
Show HN: Diffuji – a diffusion-powered instant camera(diffuji.com)
8 points by nathan-barry 7mo ago|5 comments
10.
Generating Shakespeare Without Neural Networks(nathan.rs)
5 points by nathan-barry 8mo ago|0 comments
11.
Recursion and Induction: A Self-Contained Course Using ACL2(acl2.org)
2 points by nathan-barry 8mo ago|0 comments
12.
Fidenae amphitheatre collapse (27 AD) – 20k dead(en.wikipedia.org)
2 points by nathan-barry 9mo ago|0 comments
13.
Show HN: Tiny Diffusion – Minimal diffusion LM in 364 lines(github.com)
2 points by nathan-barry 9mo ago|0 comments
14.
Anchored Diffusion Language Model(anchored-diffusion-llm.github.io)
2 points by nathan-barry 9mo ago|0 comments
15.
Text Diffusion Models Are Faster at Writing Code(nathan.rs)
3 points by nathan-barry 9mo ago|0 comments
16.
Inside ThunderKittens' Python Bindings(enbao.me)
2 points by nathan-barry 10mo ago|0 comments
17.
Kickstart.nvim: A minimal, single-file starting point for Neovim configuration(github.com)
1 points by nathan-barry 10mo ago|0 comments
18.
How we made AWS Trainium 17x faster (for conv1d)(charleshong3.github.io)
1 points by nathan-barry 10mo ago|0 comments
19.
by nathan-barry 10mo ago
Actually NVIDIA made one earlier this year, check out their Fast-dLLM paper
20.
by nathan-barry 10mo ago
Hey, I’ll add the MIT licenses later today!
21.
by nathan-barry 10mo ago
Yes, this is the case. During training, the model will get a sequence of text (ex, 512 tokens long) with a percentage of them masked out (with a special <MASK> token). It learns how to unmask those tokens to construct the original tex
22.
Show HN: Tiny Diffusion – A character-level text diffusion model from scratch(github.com)
172 points by nathan-barry 10mo ago|20 comments
23.
The Manuscripts of Edsger W. Dijkstra(cs.utexas.edu)
268 points by nathan-barry 10mo ago|125 comments
24.
Streaming DiLoCo: Towards a Distributed Free Lunch(arxiv.org)
2 points by nathan-barry 11mo ago|0 comments
25.
Fast-DLLM: Training-Free Acceleration of Diffusion LLM(arxiv.org)
70 points by nathan-barry 11mo ago|4 comments
26.
BERT is just a single text diffusion step(nathan.rs)
455 points by nathan-barry 11mo ago|110 comments
27.
Rustree: Visualize Rust ASTs in the Browser(rustree.snprajwal.com)
6 points by nathan-barry 1y ago|1 comments
28.
Diffusion is not necessarily Spectral Autoregression(fabianfalck.com)
3 points by nathan-barry 1y ago|0 comments
29.
by nathan-barry 1y ago
HTML version: https://www.cs.utexas.edu/~EWD/transcriptions/EWD13xx/EWD130...
30.
The End of Computing Science (2000) – Edsger W. Dijkstra [pdf](cs.utexas.edu)
1 points by nathan-barry 1y ago|1 comments

More ›