INDEX

Explanations

references to internet trends and viral content (memes)

oai_token-act-pair · gpt-3.5-turbo

mentions of memes and discussions surrounding their popularity

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

Interstitial

-0.73

Jewish

-0.72

idden

-0.69

 charism

-0.68

×ĳ

-0.67

OTT

-0.65

obar

-0.65

 Pall

-0.64

enting

-0.63

 earthly

-0.63

POSITIVE LOGITS

 meme

1.20

 memes

1.15

pty

0.83

eers

0.81

lore

0.81

ĸļ

0.79

ery

0.75

bris

0.71

achine

0.70

eer

0.70

Activations Density 0.005%

references to internet trends and viral content (memes)

mentions of memes and discussions surrounding their popularity

No Comments

No Known Activations