INDEX

Explanations

identifying statements and problems

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 évidemment

1.41

 eminently

1.37

 dédi

1.29

 désormais

1.27

 hervorrag

1.25

 весьма

1.23

 durchaus

1.22

 belki

1.22

 буквально

1.21

 absolutamente

1.20

POSITIVE LOGITS

 __________

2.22

 ____________

2.22

 ________

2.21

 _____

2.20

 _____________

2.20

_________

2.16

 _______

2.14

___________

2.10

_____

2.10

 ______

2.10

Activations Density 0.228%