INDEX

Explanations

celebrity names, particularly Kanye West

oai_token-act-pair · gpt-3.5-turbo

prominent names in the entertainment industry, especially musicians and celebrities

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

ugu

-0.81

thia

-0.72

ajor

-0.72

ulhu

-0.71

Assembly

-0.70

ettlement

-0.70

INESS

-0.69

dayName

-0.68

acle

-0.68

atchewan

-0.67

POSITIVE LOGITS

 Kanye

0.92

 Kardashian

0.86

cÃ©

0.82

ipedia

0.73

stress

0.72

weather

0.70

efully

0.67

Bey

0.66

anye

0.66

 West

0.65

Activations Density 0.022%

celebrity names, particularly Kanye West

prominent names in the entertainment industry, especially musicians and celebrities

No Comments

No Known Activations