INDEX

Explanations

abbreviations

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 care

-0.60

care

-0.52

 cares

-0.49

 Care

-0.47

 needs

-0.47

 cared

-0.46

rinos

-0.44

needs

-0.44

an

-0.43

 leaving

-0.42

POSITIVE LOGITS

#+#

0.81

rungsseite

0.77

 defaultstate

0.77

 فريبيس

0.76

دانشنامهٔ

0.70

 igång

0.68

oa̍t

0.68

0.61

WriteTagHelper

0.59

 linkovi

0.58

Activations Density 0.010%