INDEX

Explanations

past tense verbs followed by nouns

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

-1.86

 entire

-1.55

 llamado

-1.45

 allow

-1.38

 nazy

-1.28

 then

-1.22

</u>

-1.20

 ДЛЯ

-1.20

 juzgar

-1.20

 large

-1.19

POSITIVE LOGITS

を使用

1.77

焢

1.37

kunjung

1.36

禤

1.35

 alho

1.34

 here

1.34

真実

1.30

ブログ村

1.27

 ſch

1.26

 haki

1.26

Activations Density 0.051%