INDEX

Explanations

mentions of political figures, specifically prime ministers

oai_token-act-pair · gpt-3.5-turbo

mentions of political leaders, specifically prime ministers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 Canaver

-0.84

 Soup

-0.83

 Humanity

-0.78

Kid

-0.77

 Syndrome

-0.76

 Schwar

-0.76

 Stick

-0.76

 Crossing

-0.76

 Tribe

-0.72

 Volunteers

-0.71

POSITIVE LOGITS

 minister

1.31

 ministerial

1.22

 ministers

1.08

val

1.05

 directive

0.91

 contractor

0.87

knit

0.85

 suspect

0.81

rib

0.79

mo

0.79

Activations Density 0.016%

mentions of political figures, specifically prime ministers

mentions of political leaders, specifically prime ministers

No Comments

No Known Activations