INDEX

Explanations

mentions of using or applying a brush

oai_token-act-pair · gpt-3.5-turbo

references to various uses and types of brushes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

uthor

-0.72

 Defenders

-0.70

reme

-0.69

igmat

-0.67

 Mandela

-0.65

verson

-0.65

 Founding

-0.65

 regress

-0.64

infeld

-0.63

Sax

-0.62

POSITIVE LOGITS

brush

0.95

stroke

0.94

tip

0.85

stro

0.85

sonian

0.85

 brush

0.83

heet

0.81

 strokes

0.81

chery

0.81

 brist

0.81

Activations Density 0.053%

mentions of using or applying a brush

references to various uses and types of brushes

No Comments

No Known Activations