INDEX

Explanations

Code snippets

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

 purpoſe

-0.90

 pinulongan

-0.84

 Theſe

-0.82

 Jefus

-0.82

WriteTagHelper

-0.82

 Majefty

-0.82

 uſe

-0.79

 pleaſure

-0.79

 Monfieur

-0.76

 متعلقه

-0.75

POSITIVE LOGITS

wij

0.49

Olig

0.47

PropertyChanging

0.46

 इस्त

0.46

ameran

0.43

이션

0.43

 Hyman

0.43

ante

0.42

逝

0.42

 те

0.42

Activations Density 1.698%