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

baum

-0.80

rand

-0.74

ovo

-0.70

este

-0.69

gel

-0.69

span

-0.68

ton

-0.64

chev

-0.64

Ãł

-0.64

jong

-0.63

POSITIVE LOGITS

 aforementioned

0.91

 latter

0.90

oret

0.87

 extent

0.84

sembly

0.83

 remainder

0.82

ories

0.82

 fundamentals

0.79

 resultant

0.78

 Rockies

0.76

Activations Density 0.000%

No Known Activations

This feature has no known activations.