INDEX

Explanations

creator or owner identifier

This neuron detects property or field names in code/JSON that label entities or roles (e.g. “creator,” “owner,” “sender,” “from,” “to,” “artistName,” “DeviceId,” etc.).

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

。

-1.14

etc

-1.08

↵↵

-1.07

 myriad

-1.07

чика

-1.06



-1.03

 որ

-1.03

-1.02

“

-1.02

-1.00

POSITIVE LOGITS

who

1.96

BIDDEN

1.29

だからこそ

1.27

 gimana

1.21

 especta

1.20

tragung

1.20

 âgées

1.16

避難

1.15

Ꮸ

1.14

 manualidades

1.11

Activations Density 0.009%