INDEX

Explanations

references related to cooking or cookbooks

oai_token-act-pair · gpt-3.5-turbo

mentions of cooking

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

 Citizens

-0.79

hovah

-0.68

emale

-0.67

posure

-0.65

 guiActiveUn

-0.64

 Civic

-0.62

ilitary

-0.62

 Territories

-0.62

 Rockefeller

-0.61

innie

-0.61

POSITIVE LOGITS

 cook

1.28

books

1.18

 cooks

1.14

cook

1.08

book

1.06

washer

1.06

Cook

1.00

ware

0.91

 cooked

0.89

fry

0.87

Activations Density 0.008%

references related to cooking or cookbooks

mentions of cooking

No Comments

No Known Activations