INDEX

Explanations

content published or shared online, including articles, videos, photos, and social media posts

oai_token-act-pair · gpt-3.5-turbo

instances of publications or reports related to media or news articles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Flavoring

-0.72

aea

-0.71

ajor

-0.71

imum

-0.70

hei

-0.69

ensis

-0.67

ipolar

-0.67

thood

-0.66

estro

-0.66

paces

-0.66

POSITIVE LOGITS

 photographs

1.40

 tweets

1.40

 screenshots

1.38

 photos

1.37

 pictures

1.36

 memes

1.35

 videos

1.34

 messages

1.34

 excerpts

1.31

 images

1.30

Activations Density 0.357%

content published or shared online, including articles, videos, photos, and social media posts

instances of publications or reports related to media or news articles

No Comments

No Known Activations