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

ingen

-0.75

tox

-0.69

iii

-0.67

char

-0.65

otent

-0.64

udic

-0.64

vind

-0.64

igne

-0.62

ii

-0.62

dam

-0.61

POSITIVE LOGITS

 range

1.81

 Range

1.39

 ranges

1.18

range

1.11

 relative

0.89

 Walters

0.81

 lower

0.78

 Lower

0.76

Range

0.75

orsi

0.74

Activations Density 0.000%

No Known Activations

This feature has no known activations.