Guiferviz

I love neural networks and I think they love me back 😍

[Guiferviz @guiferviz]

While people are doing mind-blowing stuff with #GoogleGemini, I’m just getting the same image three times. Hard to believe we’re using the same model.
Top left: original.
Top right: supposedly photorealistic.
Bottom left: Disney 2D Mickey Mouse style
Bottom right: Pixar-style 3D.

[Guiferviz @guiferviz]

Adding columns is harmless… until PySpark refuses to move.
What really happens when you try summing 40 of them?
https://t.co/snYJs5MttT
#pyspark #spark #bigdata #dataengineering

[marimo @marimo_io]

🌟 Spotlight on @guiferviz!

How cell choices skew your maps.

This blog post brings the Modifiable Areal Unit Problem to life — tweak grid type, cell size and orientation to see choropleth patterns shift in real time.

Read → https://t.co/IpCEltDC83

[Denis @tg_bomze]

Face Depixelizer

Given a low-resolution input image, model generates high-resolution images that are perceptually realistic and downscale correctly.

😺GitHub: https://t.co/0WBxkyWkiK
📙Colab: https://t.co/q9SIm4ha5p

P.S. Colab is based on the
https://t.co/fvEvXKvWk2

[Andrej Karpathy @karpathy]

One (of many) amusing socially awkward remote conference call interactions is when the speaker makes a joke. Everyone is mutated so it seems like it awkwardly falls flat, and noone wants to unmute just to say "haha". Calls still need many more features to bridge the real life gap

[François Chollet @fchollet]

Looks useful: Keras-OCR (3rd-party package) provides out-of-the-box OCR models and an end-to-end training pipeline to build new OCR models

https://t.co/q7rN49VjvD

[Well-Being in ML @WellBeingInML]

The only universal HOW TO holiday
+ more of the things you enjoy
- less of the things you dislike

[Xander Steenbrugge @xsteenbrugge]

"Self-Supervised GANs via Auxiliary Rotation Loss"

A simple trick that yields better image representations and impressive improvements for training unsupervised GANs on class conditional data!

Arxiv: https://t.co/42Oq6e13Yz
GitHub: https://t.co/t5jRKenV4a

[Andy Coenen @_coenen]

Understanding UMAP - a high-level introduction to how the algorithm works, how to use it effectively, and how it compares with t-SNE.

https://t.co/Dt9Ap9G2Mc

[Gary Marcus @GaryMarcus]

Since @OpenAI still has not changed misleading blog post about "solving the Rubik's cube", I attach detailed analysis, comparing what they say and imply with what they actually did. IMHO most would not be obvious to nonexperts.

Please zoom in to read & judge for yourself.

[Dmytrii S @ml_review]

Mish: Self Regularized Non-Monotonic Activation Function
By @DigantaMisra1
𝑓(𝑥)=𝑥⋅𝑡𝑎𝑛ℎ(𝑠𝑜𝑓𝑡𝑝𝑙𝑢𝑠(𝑥))
Increased accuracy over Swish/ReLU
Increased performance over Swish

Github
https://t.co/RYzuj0xhDN
ArXiv
https://t.co/YJKTd4yKvr

[Mark Riedl @mark_riedl]

Today’s realities:

1. Non-neural net code written more than 5 years ago works just fine

2. Neural net code written a year ago on TensorFlow doesn’t work anymore

[𝙷𝚒𝚖𝚊 𝙻𝚊𝚔𝚔𝚊𝚛𝚊𝚓𝚞 @hima_lakkaraju]

Excited to announce that I am teaching a *new* course on ML interpretability and explainability at Harvard this semester. Course webpage: https://t.co/jXNJjYuWy7. Lecture slides and background materials are being uploaded as the class progresses. Lecture videos coming soon!

[arxiv @arxiv_org]

The Theory Behind Overfitting, Cross Validation, Regularization, Bagging, and Boosting: T... https://t.co/J5gPgPXOMp

[Google DeepMind @GoogleDeepMind]

Today we are excited to release video recordings of lectures from "Advanced Deep Learning and Reinforcement Learning", a course on deep RL taught at @UCL earlier this year by DeepMind researchers:

https://t.co/znsWtTxQcN

Enjoy!

[Tracy Chou 🌻 @triketora]

Are Pop Lyrics Getting More Repetitive? https://t.co/6GK6WKGm96 this is an extremely well executed article

[Nick Frosst @nickfrosst]

1/7 Our new paper on adversarial attack detection and capsule networks with @sabour_sara and Geoff Hinton is out on arxiv today! https://t.co/eiKqc2FcKM it will be presented at the #NeurIPS Workshop on Security. Don't have time to read the paper? Read this thread instead! :)

[Ofir Press @OfirPress]

Love is all you need, and attention may not be all you need! We show that a simple, attentionless translation model that uses a constant amount of memory performs on par with the Bahdanau attention model. https://t.co/xeKB8NBl7a with @nlpnoah

[Andy Brock @ajmooch]

Large-Scale GAN Training: My internship project with Jeff and Karen.

We push the SOTA Inception Score from 52 -> 166+ and give GANs the ability to trade sample variety and fidelity.

https://t.co/Vi1CMP2ZhX