INDEX

Explanations

words related to transformation or change

oai_token-act-pair · gpt-3.5-turbo

unqualified cautionary statements or warnings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Pod

-0.70

ell

-0.67

charg

-0.65

oris

-0.64

 QuÃ©

-0.64

 Thou

-0.62

});

-0.61

 Finder

-0.60

 Trance

-0.59

notations

-0.59

POSITIVE LOGITS

krit

0.85

minist

0.72

undown

0.69

jri

0.69

 Augusta

0.68

ä¸Ń

0.65

eca

0.65

mund

0.64

hung

0.64

emon

0.62

Activations Density 0.000%

words related to transformation or change

unqualified cautionary statements or warnings

No Comments

No Known Activations