INDEX

Explanations

when verb is used

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

 cordones

-1.64

踭

-1.40

 écrans

-1.32

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.30

 centaines

-1.30

 économies

-1.29

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.27

 écran

-1.25

鵙

-1.24

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-1.23

POSITIVE LOGITS

in

2.31

on

2.05

for

1.74

 from

1.67

to

1.65

 with

1.48

 before

1.45

at

1.33

 during

1.30

by

1.21

Activations Density 0.035%