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

chev

-0.84

oka

-0.82

ovo

-0.78

 admire

-0.75

pard

-0.74

ĺħ

-0.74

uti

-0.74

cro

-0.73

apon

-0.73

ardon

-0.72

POSITIVE LOGITS

 Anonymous

0.91

Anonymous

0.72

 Strait

0.68

WC

0.66

 Countries

0.65

 Spear

0.63

 Crus

0.62

 revolutions

0.60

 myths

0.60

 Memories

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.