INDEX

Explanations

whatever followed by a noun or pronoun

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-3.52

’

-2.88

 ngan

-2.72

臜

-2.66

 教室

-2.66



-2.55

archiwizowane

-2.53

蔸

-2.52

-2.48



-2.47

POSITIVE LOGITS

・「

2.69

 Many

2.48

we

2.39

 multitude

2.33

 Some

2.33

 Surprisingly

2.28

 моя

2.22

 Numerous

2.20

чины

2.14

 That

2.13

Activations Density 0.008%