INDEX

Explanations

there exists

The neuron activates on numeric tokens—especially numbers (including integers and decimals) appearing in the text.

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

 ritmo

-1.38

 melap

-1.31

 merciless

-1.31

 amado

-1.30

 塗装

-1.30

 erneu

-1.30

 音乐

-1.30

 苹果

-1.29

 机器

-1.29

POSITIVE LOGITS

 necesarias

1.59

or

1.58

 резулта

1.38

and

1.34

mId

1.32

 vindt

1.31

 prachtig

1.30

 중요

1.30

 geweldig

1.29

↵

1.27

Activations Density 0.058%