INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

sing

-0.81

enei

-0.77

ighting

-0.76

mite

-0.74

BuyableInstoreAndOnline

-0.73

smoking

-0.70

atcher

-0.70

Writer

-0.70

oing

-0.69

rontal

-0.69

POSITIVE LOGITS

 Tsarnaev

0.79

 Serge

0.78

 Russ

0.72

rett

0.71

Nev

0.70

————

0.70

 Uzbek

0.68

 Alexandra

0.67

 Varg

0.67

Taj

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.