INDEX

Explanations

phrases related to transformation or change

oai_token-act-pair · gpt-3.5-turbo

instances of the word "turn" and its variants, indicating transformation or change

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

icipated

-0.73

accompan

-0.72

è¦ļéĨĴ

-0.67

rising

-0.63

ptoms

-0.62

hess

-0.62

enegger

-0.62

ITNESS

-0.57

erest

-0.55

itely

-0.55

POSITIVE LOGITS

 into

1.39

into

1.24

 upside

1.23

 INTO

1.19

 Into

1.13

 sideways

0.94

 onto

0.93

 inward

0.92

 overboard

0.89

 loose

0.81

Activations Density 0.093%

phrases related to transformation or change

instances of the word "turn" and its variants, indicating transformation or change

No Comments

No Known Activations