INDEX

Explanations

occurrences of the word "Gothic" with varying strengths of activation

oai_token-act-pair · gpt-3.5-turbo

references to Gothic architecture

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

 Kislyak

-0.72

 Wiggins

-0.68

assetsadobe

-0.67

 Cheong

-0.66

andum

-0.66

mble

-0.66

 Waterloo

-0.65

bles

-0.64

lished

-0.64

XT

-0.63

POSITIVE LOGITS

ouse

0.89

ic

0.76

iop

0.74

child

0.74

face

0.72

 diapers

0.72

bound

0.71

some

0.69

alos

0.69

aus

0.69

Activations Density 0.011%

occurrences of the word "Gothic" with varying strengths of activation

references to Gothic architecture

No Comments

No Known Activations