INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

undy

-0.65

sonian

-0.64

 wonder

-0.62

 Noir

-0.61

 Rouge

-0.60

 rooft

-0.60

 baker

-0.58

fav

-0.57

owitz

-0.57

 Math

-0.57

POSITIVE LOGITS

emort

0.81

Sym

0.77

ptoms

0.73

Lear

0.70

 Rebell

0.69

 Canaver

0.68

omaly

0.68

 tsun

0.67

RH

0.67

apore

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.