INDEX

Explanations

only show or disallow

The neuron fires on words and phrases expressing permission or obligation (e.g. “mean,” “required,” “allowed,” “you are to…”).

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

お疲れ様でした

-1.30

 musí

-1.28

 desplaz

-1.23

 debía

-1.20

 pomá

-1.20

おわりに

-1.18

を作って

-1.15

 podía

-1.15

 required

-1.13

 another

-1.11

POSITIVE LOGITS

 ($('#

1.23

{?>

1.16

자인

1.15

 ($("#

1.13

今は

1.05

as

1.05

 includes

1.03

 حتی

1.02

值为

1.00

Activations Density 0.195%