INDEX

Explanations

philosophers and thinkers

This neuron activates on tokens that are part of proper names—especially names of authors, scholars, or other capitalized entities.

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

ૉ

-1.33

 serif

-1.12

 jantung

-1.09

马克

-0.98

its

-0.96

壁纸

-0.96

ادث

-0.93

 Arter

-0.91

ária

-0.91

 hasn

-0.89

POSITIVE LOGITS

 himself

1.98

 نفسه

1.52

himself

1.45

 sám

1.21

 cérami

1.19

 kabát

1.14

 lèvres

1.13

等人

1.11

 himſelf

1.07

 granat

1.06

Activations Density 0.034%