INDEX

Explanations

after, then, and when

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

 ocasion

-1.43

ectores

-1.41

 الشاشة

-1.38

跻

-1.33

 десер

-1.31

 fredag

-1.30

tbx

-1.30

惩

-1.30

 verhind

-1.29

を中心

-1.29

POSITIVE LOGITS

 began

1.67

 after

1.59

 continued

1.54

by

1.51

 then

1.48

two

1.41

 subsequently

1.35

 soon

1.34

but

1.32

 shortly

1.32

Activations Density 0.015%