INDEX

Explanations

keywords related to different countries in Europe

oai_token-act-pair · gpt-3.5-turbo

references to Europe

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

abal

-0.77

SourceFile

-0.75

eful

-0.73

wx

-0.68

awan

-0.67

ruff

-0.66

rences

-0.65

 remem

-0.65

aroo

-0.64

washer

-0.64

POSITIVE LOGITS

 Europe

3.53

Europe

2.99

 European

2.26

 Europeans

2.26

European

2.13

 Germany

2.06

 France

2.03

Germany

1.95

 Italy

1.93

 Scandinav

1.93

Activations Density 0.018%

keywords related to different countries in Europe

references to Europe

No Comments

No Known Activations