INDEX

Explanations

references and allusions

This neuron detects words and phrases that signal a reference or tribute (e.g. “nod to,” “alludes to,” “tribute to,” “homage to”).

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

 there

-2.17

There

-1.70

 this

-1.55

this

-1.37

これを

-1.34

 convencer

-1.31

 konnte

-1.30

 этот

-1.29

-1.28

</h2>

-1.27

POSITIVE LOGITS

to

1.58

archiwizowane

1.44

amerikanischer

1.42

colgante

1.38

rennen

1.38

wachung

1.38

KIM

1.37

anud

1.36

 begge

1.35

 Meksi

1.34

Activations Density 0.040%