INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ylum

-0.85

urgical

-0.79

hig

-0.79

ucer

-0.79

hew

-0.76

ewitness

-0.74

ologically

-0.73

advert

-0.72

estation

-0.71

ometimes

-0.71

POSITIVE LOGITS

KC

0.71

 Sterling

0.69

 Studios

0.68

BCC

0.68

Wiz

0.68

AV

0.66

 Silver

0.66

KC

0.65

($)

0.64

Moj

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.