INDEX

Explanations

alignment and justification CSS/HTML attributes

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

-1.64

by

-1.60

内外

-1.59

 adanya

-1.55

 have

-1.55

be

-1.52

}$.

-1.47

平衡

-1.46

="

-1.45

」。

-1.45

POSITIVE LOGITS

2.02

Lovely

1.52

陆续

1.47

後來

1.38

今年の

1.36

lepaskan

1.32

降临

1.32

子供の

1.31

 epitome

1.31

の通り

1.30

Activations Density 0.001%