INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

)</

-0.70

swick

-0.68

âĢ¦)

-0.68

EStream

-0.68

 bargain

-0.68

esides

-0.66

":""},{"

-0.65

veland

-0.65

selves

-0.64

POSITIVE LOGITS

idium

0.78

Mek

0.67

 horns

0.66

 Jericho

0.66

 Sonny

0.65

omics

0.65

Ali

0.63

 Jacobs

0.62

mone

0.62

Jub

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.