INDEX

Explanations

understanding user's intent

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

¿?

1.39

 metà

1.29

½

1.24

è

1.22

 funziona

1.21

 dunno

1.20

 sits

1.16

!!.

1.14

¼

1.13

 está

1.13

POSITIVE LOGITS

how

1.56

how

1.51

 كيفية

1.41

如何

1.34

 faptul

1.22

 bagaimana

1.19

如何在

1.17

是否存在

1.16

 எவ்வாறு

1.14

 bahwa

1.14

Activations Density 1.077%