Crossposted from the AI Alignment Forum. May contain more technical jargon than usual.

As more people begin work on interpretability projects which incorporate dictionary learning, it will be valuable to have high-quality dictionaries publicly available.[1] To get the ball rolling on this, my collaborator (Aaron Mueller) and I are:

open-sourcing a number of sparse autoencoder dictionaries trained on Pythia-70m MLPs
releasing our repository for training these dictionaries.

Let's discuss the dictionaries first, and then the repo.

The dictionaries

The dictionaries can be downloaded from here. See the sections "Downloading our open-source dictionaries" and "Using trained dictionaries" here for information about how to download and use them. If you use these dictionaries in a published paper, we ask that you mention us in the acknowledgements.

We're releasing two sets of dictionaries for EleutherAI's 6-layer pythia-70m-deduped model. The dictionaries in both sets were trained on 512-dimensional MLP output activations (not the MLP hidden [...]

---

Outline:

(00:41) The dictionaries

(05:30) The dictionary learning repository

The original text contained 1 footnote which was omitted from this narration.

---

First published:
December 5th, 2023

Source:
https://www.lesswrong.com/posts/AaoWLcmpY3LKvtdyq/some-open-source-dictionaries-and-dictionary-learning

---

Narrated by TYPE III AUDIO.

“Some open-source dictionaries and dictionary learning infrastructure” by Sam Marks

LessWrong (30+ Karma)