INDEX

Explanations

character names

The neuron is triggered by capitalized named entities—proper names of characters, products, places, etc.

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

 السياس

-1.58

лкой

-1.42

ớm

-1.41

ptons

-1.38

ּוֹ

-1.37

 gatos

-1.28

 pisa

-1.27

έ

-1.27

 domingos

-1.25

 internacional

-1.24

POSITIVE LOGITS

who

1.57

"}";

1.30

 beijo

1.30

 sofrer

1.21

 braço

1.19

 desejo

1.19

 religieuses

1.19

 ネタバレ

1.19

 lembrar

1.18

 قوات

1.16

Activations Density 0.013%