INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

hower

-0.80

ecause

-0.78

iqueness

-0.73

ournal

-0.70

cffff

-0.70

 stren

-0.68

Gin

-0.68

anqu

-0.68

hovah

-0.68

isure

-0.67

POSITIVE LOGITS

ney

0.70

wolf

0.69

 animate

0.65

NEY

0.65

marks

0.63

net

0.63

 Urban

0.63

 Shutterstock

0.63

fed

0.62

 translate

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.