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

aden

-0.73

 unden

-0.73

Ns

-0.72

agy

-0.70

growth

-0.68

aney

-0.67

ebin

-0.65

andel

-0.65

aken

-0.64

odies

-0.64

POSITIVE LOGITS

âĢ

1.03

 âĢ

0.85

âĺ

0.80

 Clement

0.75

ouver

0.72

esville

0.67

âĢł

0.66

 Tobias

0.66

 includ

0.64

wip

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.