INDEX

Explanations

possessive across languages

The neuron flags specialized terminology—brand or model names, materials, and technical spec tokens (e.g. “Talalay,” “Dunlop,” “3 G,” “LTE,” “4 G”).

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

 thậm

-1.34

sibles

-1.32

甚至

-1.28

 needed

-1.24

 segni

-1.22

 близь

-1.20

iredo

-1.20

artement

-1.20

inine

-1.20

 отвечает

-1.20

POSITIVE LOGITS

you

1.80

 your

1.78

or

1.66

 شما

1.54

 traktor

1.52

 आपके

1.45

 вам

1.30

 сахара

1.27

 kategor

1.23

 apartament

1.23

Activations Density 0.050%