Datasets:
Tasks:
Text Generation
Modalities:
Text
Languages:
English
Size:
1K - 10K
ArXiv:
Tags:
sparse-autoencoder
mechanistic-interpretability
representation-verbalization
feature-explanation
neuronpedia
License:
Commit History
-
Add 1B-to-27B adapter evaluation data
Verified signature3f87961
-
Updata meta data
Verified signature02924d4
-
docs: add code repository link
Verified signature7d1c305 -
Upload README.md with huggingface_hub
Verified signature267cecb -
Upload README.md with huggingface_hub
Verified signaturedd67f2d -
Organize evaluation sets under evaluation directory
Verified signaturebc1c687 -
Upload README.md with huggingface_hub
Verified signature07b85de -
Upload evaluation/global_gold_1000.json with huggingface_hub
Verified signaturee351507 -
Upload evaluation/low_index_gold_200.json with huggingface_hub
Verified signature6514de3 -
Upload evaluation/global_train_standard_1000.json with huggingface_hub
Verified signatureb0030b7 -
Upload SAEVerbalizer evaluation sets
Verified signature40a0d25 -
initial commit
Verified signature2fd4eca