INDEX

Explanations

Entity names followed by parenthetical descriptions

This neuron detects the capitalized preposition “In” at the start of a sentence or section.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 that

-1.65

of

-1.52

 their

-1.31

 those

-1.23

get

-1.20

取り組

-1.15

到你

-1.12



-1.08

</b>

-1.05

Тогда

-1.04

POSITIVE LOGITS

 cabelos

1.38

vantaged

1.30

cedent

1.26

doge

1.14

 GOING

1.13

 surtido

1.13

igert

1.12

 ווע

1.12

ﮟ

1.11

 meninos

1.09

Activations Density 0.029%