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

Negative Logits

apego

-0.89

ivas

-0.75

uries

-0.74

requ

-0.74

hyde

-0.73

ĨĴ

-0.71

aji

-0.71

efficients

-0.68

urt

-0.67

vati

-0.67

POSITIVE LOGITS

 throats

0.69

 ours

0.68

cloth

0.67

oration

0.66

onom

0.63

pper

0.62

 checked

0.62

TN

0.62

bands

0.62

DID

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.