Every premium and members-only resource for $9.99/month, as a no-commitment subscription. See the plans

PalateKit

Sign in
Menu

Request a resource

Describe the resource you are missing in one sentence. We will email you when it is published.

0 / 280

A page, a video or an image online.

Unlock PalateKit

Unlock everything

Plan

Paying means accepting our terms of sale.

Secure payment by Stripe. Receipts and invoices sent by email.

Welcome back

Sign in to install, download or save this resource. A free account is enough for free resources.

Sparse autoencoders to interpret a model v1.0.0

Use this resource to train and analyse sparse autoencoders to interpret a model's features

New

  • 377 downloads
  • 102 likes
  • 122 saves

Free

Install with the CLI

In a terminal, from the folder you want:

npx --yes palatekit@latest install sparse-autoencoder-interpretability-2

With no option, the skill goes to ~/.agents/skills/; add --agent claude-code (or codex, cursor…) to target an agent.

First time: connect the CLI with

npx --yes palatekit@latest login

The agent loads a model and a pre-trained autoencoder with SAELens, examines the features activated by texts, then can train a new autoencoder. You get an analysis of a model's internal features, for research use.

Mostly picked by

Updated November 2, 2026

Card for the “Sparse autoencoders to interpret a model” skill

Details

Works in
Install guides
Generates with
Your agent's model
Licence
MIT

Similar resources

View all

What you get

  • SKILL.md: concepts, installation and loading a pre-trained autoencoder
  • references/: API reference, README and step-by-step tutorials

Good fit for

  • You study a model's polysemantic neurons
  • You want to identify an LLM's internal features in a research project

Not for

    Tags

    • interpretability
    • autoencoder
    • llm
    • research
    • features

    About

    You say

    “Load this model's autoencoder and list the features most activated by this text.”

    Card for the “Sparse autoencoders to interpret a model” skill

    Skill published by Orchestra Research under the MIT license, delivered in its original version.

    Show more

    How to use it

    1. Install the skill in your agent's skills folder
    2. Ask your agent to train and analyse sparse autoencoders to interpret a model's features
    3. Review and adjust the result

    Content preview

    Files

    • LICENSE1 KB
    • NOTICE-PALATEKIT.md0.5 KB
    • references/api.md6.8 KB
    • references/README.md2 KB
    • references/tutorials.md9.2 KB
    • SKILL.md14.7 KB

    Origin and licence

    Listed unmodified

    Author
    Orchestra Research
    Source
    github.com/nousresearch/hermes-agent/tree/0c74cc682dc81bf78b78a1737b3ee77a820ccb79/optional-skills/mlops/saelens
    Licence
    MIT

    Notice

    « saelens » (Orchestra Research, MIT) livré dans sa version d'origine : fiche en français et en anglais par PalateKit.

    User reviews

    –

    No ratings yet

    Loading reviews…

    Delete your review?

    Your rating and comment will be removed from the page.