INDEX

Explanations

understandably wary or skeptical

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

injury

1.43

petition

1.42

and

1.40

advers

1.39

ь

1.38

sep

1.35

Gb

1.34

capture

1.34

λικ

1.32

this

1.32

POSITIVE LOGITS

 geralmente

1.72

 certamente

1.62

િજા

1.61

 souvent

1.60

 básicamente

1.58

 skoro

1.57

やや

1.55

 pár

1.54

 buscador

1.54

 élément

1.54

Activations Density 0.095%