r3ndd

How is the AI Dungeon able to generate text so quickly using the 1558M model in Collaboratory?

15

A project called the AI Dungeon (https://colab.research.google.com/github/nickwalton/AIDungeon/blob/master/AIDungeon_2.ipynb) supposedly uses the 1558M model to generate stories in Collaboratory. However, this strikes me as odd because 1) GPT-2 Simple can't load the...

Clients do not know when children have been filtered

1

Right now clients cannot distinguish between a child of an array/map being filtered and a child simply not being updated. While this may be desirable for certain applications, it is...

Changed supervised PCA to use center of contrasting vectors

4

This is a fix in extract.py for #28 I also re-ran the experiments.ipynb notebook to see the outputs. As expected, using this method seems to create stronger vectors, such that...

In the original repeng paper they mostly described the unsupervised version of PCA, where they randomly paired the hidden state vectors and performed PCA on these pairs. This is fundamentally...

Help with LSTM basics in synaptic.js

7

I looked through the Wikipedia Reading example for help, but I couldn't actually get the example to run and the code itself is hard to follow as someone who hasn't...

Bacterial Contamination and Infections

1

This is a plan to add simple bacterial contamination to the station, separate from but compatible with and complimenting some proposed virology systems such as #332 . All percentages/probabilities I...