INDEX

Explanations

mentions of the word "slug."

oai_token-act-pair · gpt-3.5-turbo

mentions of slugs or slug-related terms

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

mental

-0.72

 Scotia

-0.72

raising

-0.64

lesiastical

-0.64

 Schools

-0.63

foreseen

-0.63

SPONSORED

-0.63

SOURCE

-0.63

Geo

-0.62

 Passage

-0.62

POSITIVE LOGITS

 slug

1.38

 Slug

1.22

gers

0.88

bats

0.85

zag

0.84

gery

0.83

ging

0.81

bowl

0.80

ger

0.76

swick

0.75

Activations Density 0.008%

mentions of the word "slug."

mentions of slugs or slug-related terms

No Comments

No Known Activations