INDEX

Explanations

punctuation followed by URLs or citations

The neuron activates on numeric tokens (digits or numbers), i.e. it flags standalone numbers and numeric references.

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

 Webseiten

-0.92

\'

-0.91

hti

-0.91

 przede

-0.90

Ｖ

-0.89

нциклопе

-0.87

製の

-0.87

 Usaha

-0.86

maining

-0.86

 Erfüllung

-0.85

POSITIVE LOGITS

rased

1.11

 elated

1.10

yayım

1.09

توضیحات

1.09

 ibid

1.08

 スイッチ

1.07

Már

1.06

Ár

1.05

 definizione

1.05

ઃ

1.04

Activations Density 0.028%