INDEX

Explanations

foreign language words

The neuron preferentially lights up on non‐stop-word, content-bearing technical terms (i.e. substantive nouns and verbs) rather than on function words.

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

емъ

-1.28

 гита

-1.25

ִּ

-1.23

 دے

-1.22

といけない

-1.21

 typelib

-1.20

 objekt

-1.17

ֶּ

-1.17

</h4>

-1.17

 creado

-1.16

POSITIVE LOGITS

 Especially

1.40

 einerseits

1.24

 religieux

1.21

事实上

1.21

 insbesondere

1.21

 chré

1.20

消失了

1.18

遨

1.16

Especially

1.16

 kompetisi

1.16

Activations Density 0.427%