INDEX

Explanations

reporting verbs and research actions

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

 получить

-1.18

といった

-1.13

 تطبيق

-1.10

Únete

-1.09

 unserer

-1.08

endif

-1.07

攥

-1.06

缡

-1.05

WidgetItem

-1.05

א

-1.02

POSITIVE LOGITS

 that

2.23

its

1.52

the

1.39

 with

1.38

 where

1.29

 their

1.26

場所は

1.26

intendo

1.23

 fordert

1.21

鐮

1.20

Activations Density 0.190%