LoreonLabsPlatform

Overview

Intelligence

Markets
Builders
Research
Narratives
Ecosystems
Launchpads

Discover

Search
Sources

Jupyter Notebook

ravel

Evaluate interpretability methods on localizing and disentangling concepts in LLMs.

Jupyter NotebookEmergingcausal-interventiondisentangled-representationsinterpretabilityintervention

Stars

58

Forks

10

Contributors

1

Last push

8mo ago

Recent commits

Latest commits.

Merge pull request #3 from explanare/demos
e421fc9Jing Huang8mo ago
Update README.
4c5330fJing Huang8mo ago
Create LICENSE
5b3bbdaJing Huang21mo ago
Add the benchmark script for the google/gemma-2-2b model.
c9b3398Jing Huang21mo ago
Create a demo directory.
677e57dJing Huang21mo ago

Update README.md.

59af818Jing Huang21mo ago

Support tokenizers with a BOS token. Support multiple intervention sites.

7cd924aJing Huang21mo ago

Check training sample size.

319ab85Jing Huang21mo ago

Top contributors

Builders behind this project.