INDEX

Explanations

start of turn

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

 amid

1.41

 respectable

1.35

 affluent

1.33

 substantive

1.32

 absorb

1.27

 extravagant

1.26

 decadent

1.25

 opulent

1.25

 impregn

1.23

 film

1.21

POSITIVE LOGITS

練習

1.57

ブログ

1.52

様専用

1.50

です

1.42

报错

1.40

メソッド

1.40

修改

1.37

尝试

1.36

Github

1.35

初始化

1.35

Activations Density 0.219%