teddykoker

20 ms·

1.
by teddykoker 8mo ago
OpenEquivariance [1] is another good baseline for with kernels for the Clebsch-Gordon tensor product and convolution, and it is fully open source. Both kernel implementations have been successfully implemented into existing machine learning
2.
Nicholas Carlini – Are LLMs worth it? [video](youtube.com)
2 points by teddykoker 11mo ago|0 comments
3.
All Atom Virtual Cell(diffuse.one)
2 points by teddykoker 1y ago|0 comments
4.
Why woodpeckers don't get concussions (2024)(fermatslibrary.com)
1 points by teddykoker 2y ago|1 comments
5.
My (Not-So-Successful) Quest to Conquer the NYT Connections Game with Word2Vec(marshalljiang.com)
2 points by teddykoker 3y ago|0 comments
6.
by teddykoker 3y ago
A related line of work is "Thinking Like Transformers" [1]. They introduce a primitive programming language, RASP, which is composed of operations capable of being modeled with transformer components, and demonstrate how different
7.
Learning to Learn with Jax(teddykoker.com)
2 points by teddykoker 3y ago|0 comments
8.
Never Go Full Kelly (2021)(lesswrong.com)
7 points by teddykoker 3y ago|0 comments
9.
ClosedAI – a drop-in replacement for OpenAI with open models(github.com)
10 points by teddykoker 3y ago|0 comments
10.
Making Python faster with Rust(ohadravid.github.io)
387 points by teddykoker 3y ago|215 comments
11.
by teddykoker 4y ago
See also Radio2Speech [1] which uses a UNet to recover audio from a RF beam. [1] https://zhaorunning.github.io/Radio2Speech/
12.
by teddykoker 4y ago
Link to paper [1]. It looks like the authors construct a basic autoencoder to predict frames in videos of various videos (double pendulum, lava lamp, etc.) and then use the Levina–Bickel algorithm [2] to determine the expected "intrins
13.
by teddykoker 4y ago
According to [1], the byte pair encoding for “Apoploe vesrreaitais” (the words producing bird images) is "apo, plo, e</w>, ,ve, sr, re, ait, ais</w>", and Apo-didae & Plo-ceidae are families of birds. [1]
14.
Optimism Shapes Reality(alexw.substack.com)
1 points by teddykoker 4y ago|0 comments
15.
Learning JAX by Learning to Learn(teddykoker.com)
2 points by teddykoker 4y ago|0 comments
16.
The Reversal Test(en.wikipedia.org)
4 points by teddykoker 4y ago|0 comments
17.
Bad ML Abstractions (Generative vs. Discriminative Models)(jacobbuckman.com)
1 points by teddykoker 4y ago|1 comments
18.
Equinox: Neural networks in Jax via callable PyTrees and filtered transformation(arxiv.org)
3 points by teddykoker 4y ago|0 comments
19.
by teddykoker 4y ago
For image retrieval, have you tried using a model trained with contrastive learning (e.g. SimCLR)? This could produce better embeddings for retrieval since the model is trained to explicitly minimize euclidean distance between similar pairs
20.
by teddykoker 4y ago
I wonder how their advertised "vector database" works. kNN combined with embeddings from pre-trained deep learning models can be very useful for information retrieval, (e.g. searching for duplicate/similar images or text). In
21.
by teddykoker 4y ago
Also see Einops: https://github.com/arogozhnikov/einops , which uses a einsum-like notation for various tensor operations used in deep learning. https://einops.rocks/pytorch-examples.html shows how it c
22.
by teddykoker 4y ago
This might be what you’re looking for: https://gafniguy.github.io/4D-Facial-Avatars/
23.
by teddykoker 5y ago
Yup! Just a multinomial logistic regression model with a bunch of proprietary data and engineered features.
24.
by teddykoker 5y ago
You can find Benter’s paper about the model here: https://www.gwern.net/docs/statistics/decision/1994-benter.p... About a year ago I read this and attempted to build a similar model (with more layers ;)) usin
25.
Lightning Flash: A collection of tasks for solving problems with deep learning(github.com)
14 points by teddykoker 6y ago|1 comments
26.
DataLoaders Explained: Building a Multi-Process Data Loader from Scratch(teddykoker.com)
2 points by teddykoker 6y ago|0 comments
27.
Performers: The Kernel Trick, Random Fourier Features, and Attention(teddykoker.com)
2 points by teddykoker 6y ago|0 comments
28.
Deep Learning for Guitar Effect Emulation(teddykoker.com)
314 points by teddykoker 6y ago|159 comments