INDEX

Explanations

literal or pun

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

 such

-0.90

 както

-0.81

łych

-0.79

ujące

-0.79

done

-0.75

called

-0.75

ijn

-0.75

 known

-0.74

whatever

-0.73

ТА

-0.73

POSITIVE LOGITS

 literally

1.66

 Literally

1.62

literally

1.52

pun

1.48

 puns

1.43

 буквально

1.37

 literal

1.34

 literalmente

1.32

Literally

1.23

 pardon

1.17

Activations Density 0.035%