INDEX

Explanations

titles or headings in a text

oai_token-act-pair · gpt-3.5-turbo

references to specific subjects or entities in a list format

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

adr

-0.80

iannopoulos

-0.78

rie

-0.76

cles

-0.75

otte

-0.74

ãĥ¼ãĤ¯

-0.73

bal

-0.72

efully

-0.72

riet

-0.72

thumbnails

-0.71

POSITIVE LOGITS

wagen

0.76

 Eleven

0.74

 Away

0.72

 Waiting

0.71

 Stranger

0.70

 Flavoring

0.67

 Throw

0.67

backer

0.66

 Things

0.66

chedel

0.65

Activations Density 0.023%

titles or headings in a text

references to specific subjects or entities in a list format

No Comments

No Known Activations