INDEX

Explanations

answers, garlic bread, account

This neuron is sensitive to title‐style or heading words—capitalized tokens that appear in section or article titles.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 собрать

-1.30

 bajar

-1.29

了一个

-1.23

шир

-1.23



-1.20

了一個

-1.16

闳

-1.14

öffnet

-1.13

alamu

-1.13

rafat

-1.12

POSITIVE LOGITS

 them

1.48

₫

1.32

clave

1.30

 that

1.26

riezmann

1.17

了啊

1.16

作り方

1.15

戕

1.14

了我

1.13

)];

1.11

Activations Density 0.028%