INDEX

Explanations

the word "everywhere" occurring frequently in text segments

oai_token-act-pair · gpt-3.5-turbo

the word "everywhere."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

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

bal

-0.94

missions

-0.76

gotten

-0.71

DA

-0.70

erick

-0.68

char

-0.67

hem

-0.67

stra

-0.66

lev

-0.66

bor

-0.65

POSITIVE LOGITS

 Everywhere

1.37

 everywhere

1.11

£ı

1.09

abouts

1.00

ĸļ

0.93

 else

0.90

uria

0.83

 imaginable

0.82

Ĥİ

0.82

oun

0.79

Activations Density 0.005%

the word "everywhere" occurring frequently in text segments

the word "everywhere."

No Comments

No Known Activations