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

ļéĨĴ

-0.94

iott

-0.81

oola

-0.79

mouth

-0.77

 guiActiveUn

-0.76

ancial

-0.73

nesday

-0.72

bart

-0.68

ÃĥÃĤ

-0.66

emouth

-0.66

POSITIVE LOGITS

bird

0.74

 Americas

0.73

uana

0.68

oslav

0.67

 Oriental

0.66

leon

0.65

addons

0.63

birds

0.62

 imperialist

0.61

 narrator

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.