INDEX

Explanations

or, not A or B

The neuron detects numeric tokens and quantities (especially numbers with decimals or counts).

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

basically

-1.37

 basically

-1.29

Basically

-1.27

她和

-1.23

TacToe

-1.20

Ꮠ

-1.20

 Junio

-1.20

鈁

-1.17

řejmě

-1.16

ïde

-1.16

POSITIVE LOGITS

it

2.25

 mágico

1.38

 или

1.36

 there

1.35

on

1.34

you

1.32

 görüntüsü

1.31

了出来

1.30

瞞

1.28

ściu

1.27

Activations Density 0.034%