INDEX

Explanations

references to a spouse, specifically focusing on the word "wife."

oai_token-act-pair · gpt-3.5-turbo

references to the term "wife."

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

etting

-0.75

 Flavoring

-0.74

orescent

-0.73

obyl

-0.70

oday

-0.70

umbn

-0.68

anyl

-0.65

ums

-0.65

 constitu

-0.64

aneously

-0.63

POSITIVE LOGITS

wife

1.21

hood

0.88

cake

0.77

doctor

0.76

 wife

0.75

gdala

0.74

cook

0.74

husband

0.73

 Maria

0.73

maker

0.73

Activations Density 0.017%

references to a spouse, specifically focusing on the word "wife."

references to the term "wife."

No Comments

No Known Activations