INDEX

Explanations

common phrases and contexts

This neuron fires on text that reads like in-game narration from a tabletop RPG session—i.e. adventure log prose describing characters moving, splitting off, and entering locations.

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

 pasto

-1.58

 abstrak

-1.52

 foton

-1.51

 geene

-1.48

咂

-1.41

 skriv

-1.41

 universite

-1.40

 engel

-1.39

 tund

-1.38

 milj

-1.38

POSITIVE LOGITS

 with

1.62

 their

1.59

carian

1.39

ného

1.37

 lleno

1.33

 их

1.30

 trasero

1.27

ēju

1.23

 głowy

1.20

 dieron

1.20

Activations Density 0.015%