INDEX

Explanations

mentions or references to Australia

oai_token-act-pair · gpt-3.5-turbo

mentions of Australia

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

-0.83

ever

-0.77

lest

-0.72

:/

-0.70

angler

-0.70

{\

-0.69

ages

-0.69

aired

-0.65

Discussion

-0.64

meyer

-0.64

POSITIVE LOGITS

 Australia

1.28

Australia

1.10

 Australians

1.07

 Tasmania

1.06

NSW

0.91

yip

0.89

 Samoa

0.85

 Queensland

0.84

 Aboriginal

0.83

 Australian

0.83

Activations Density 0.013%

mentions or references to Australia

mentions of Australia

No Comments

No Known Activations