INDEX

Explanations

resistant strain, chooses to, pulled down, warned that

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

 willkommen

-1.30

 menegaskan

-1.25

annulation

-1.23

%\\

-1.19

cepatan

-1.17

ITHER

-1.17

 frequentemente

-1.15

 Baumwolle

-1.11

 rigue

-1.11

 pięk

-1.11

POSITIVE LOGITS

it

1.16

estimable

1.11

 negros

1.09

 cannot

1.06

 doesn

1.05

 extenso

1.02

 بی‌

1.01

šne

0.99

 didn

0.99

 años

0.97

Activations Density 0.042%