INDEX

Explanations

references to the country "Australia"

oai_token-act-pair · gpt-3.5-turbo

references to Australia

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

outer

-0.76

abet

-0.75

ows

-0.70

ever

-0.69

angler

-0.69

Lens

-0.69

acco

-0.68

OHN

-0.66

Drag

-0.66

uers

-0.66

POSITIVE LOGITS

 Australia

1.03

Australia

0.84

 Broadcasting

0.84

 Airlines

0.84

 Tasmania

0.83

 Australians

0.83

 Defence

0.83

 Geographic

0.79

 Bureau

0.79

 Aboriginal

0.77

Activations Density 0.015%

references to the country "Australia"

references to Australia

No Comments

No Known Activations