INDEX

Explanations

to + action/purpose

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

to

-1.63

 makes

-1.29

 being

-1.22

 gets

-1.19

 doing

-1.16

 having

-1.15

 making

-1.02

 getting

-1.02

 produces

-1.02

しくは

-0.99

POSITIVE LOGITS

 help

1.63

 umożli

1.46

 support

1.44

 complement

1.41

 accompany

1.38

ようになった

1.36

 supplement

1.31

 coincide

1.30

 zapew

1.28

 address

1.22

Activations Density 0.139%