INDEX

Explanations

terms related to Australia or Australian entities

oai_token-act-pair · gpt-3.5-turbo

references to Australia and its entities

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

 leveled

-0.75

ipation

-0.67

 leveling

-0.67

Lv

-0.66

rir

-0.65

chwitz

-0.64

 visualize

-0.63

 afterward

-0.62

 torches

-0.62

Â·Â·

-0.62

POSITIVE LOGITS

 Australian

3.73

Australian

3.48

 Australians

3.08

Australia

2.66

Austral

2.64

 Australia

2.62

 Canberra

2.31

 Queensland

2.23

 Sydney

2.16

 Aboriginal

2.12

Activations Density 0.013%

terms related to Australia or Australian entities

references to Australia and its entities

No Comments

No Known Activations