INDEX

Explanations

in phrases indicating location or inclusion

This neuron responds to short prepositions (e.g. “in,” “on,” “no,” “em”) marking location or time.

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

if

-1.54

<bos>

-1.38

 which

-1.32

 still

-1.26

rdd

-1.20

in

-1.15

 this

-1.15

isNumber

-1.05

مراجع

-1.05

şi

-1.05

POSITIVE LOGITS

ieno

1.23

φέρει

1.20

 bearbeitet

1.17

好像是

1.16

 Gesetze

1.14

 ganancias

1.13

 extraordinario

1.13

と同じ

1.13

 отсутствие

1.12

 fisuras

1.12

Activations Density 0.411%