INDEX

Explanations

nouns preceded by the word "of"

oai_token-act-pair · gpt-3.5-turbo

phrases indicating possession or belonging

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

NPR

-0.81

CEO

-0.81

 partName

-0.75

UFC

-0.72

tracking

-0.72

wcs

-0.72

kids

-0.71

pause

-0.71

 GoPro

-0.71

Bloomberg

-0.70

POSITIVE LOGITS

 antiquity

0.96

 virtue

0.89

 mankind

0.88

 Virtue

0.87

 Constantinople

0.81

 Divine

0.80

 oneself

0.80

tho

0.79

 divine

0.78

 Religion

0.76

Activations Density 0.358%

nouns preceded by the word "of"

phrases indicating possession or belonging

No Comments

No Known Activations