INDEX

Explanations

copy of

The neuron fires on occurrences of the word “copy” (especially in legal or licensing contexts, often where a document or license copy is being obtained or attached).

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

-1.61

 outras

-1.56

只剩下

-1.44

 muitos

-1.39

➨

-1.36

惴

-1.33

 başka

-1.31

いろんな

-1.31

 feront

-1.31

 muitas

-1.29

POSITIVE LOGITS

겟

1.55

»،

1.55

)،

1.47

CurtirCurtir

1.41

wwww

1.40

the

1.37

onStop

1.34

Lehr

1.33

sequently

1.31

Waga

1.31

Activations Density 0.063%