INDEX

Explanations

mentions of the word "readers"

oai_token-act-pair · gpt-3.5-turbo

references to the audience or readership of the content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asia

-0.75

 Parenthood

-0.68

 Pharma

-0.60

 Normandy

-0.60

 Kore

-0.60

WCS

-0.60

 circumstance

-0.59

 Heck

-0.59

 Guam

-0.58

 Gund

-0.58

POSITIVE LOGITS

hip

1.25

 readers

1.07

 reader

1.02

 Readers

0.95

DragonMagazine

0.94

reader

0.88

uggest

0.87

lisher

0.82

worms

0.81

boys

0.81

Activations Density 0.022%

mentions of the word "readers"

references to the audience or readership of the content

No Comments

No Known Activations