INDEX

Explanations

names and surnames

The neuron activates on capitalized word fragments that are part of proper names or titles.

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

lossians

-2.64

摭

-2.42

 serão

-2.33

趼

-2.33

 blaue

-2.31

 '+':

-2.30

dónde

-2.19

でございます

-2.17

 '*':

-2.16

辚

-2.14

POSITIVE LOGITS

其妙

2.73

螂

2.69

褔

2.47

 书法

2.44

 machte

2.39

鐿

2.38

at

2.38

2.33

 Mitarbeitern

2.20

 They

2.17

Activations Density 0.011%