INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 chall

-0.72

 horm

-0.68

iquette

-0.65

 preach

-0.64

 Browne

-0.63

 previews

-0.63

 redistributed

-0.62

 blacklist

-0.60

Rue

-0.60

 Casual

-0.60

POSITIVE LOGITS

SPONSORED

0.78

eus

0.76

Temperature

0.75

 guiActiveUnfocused

0.72

ãĤ¿

0.72

lust

0.70

Utah

0.69

0.68

BACK

0.68

TEXTURE

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.