INDEX

Explanations

states, conditions, and missing things

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ْع

1.28

order

1.26

atang

1.24

alarm

1.22

ამო

1.22

ān

1.21

 voila

1.20

एँ

1.17

atter

1.16

engar

1.16

POSITIVE LOGITS

 brown

1.44

 中央

1.38

 retir

1.35

³

1.34

 apical

1.33

 மாநில

1.32

 agricultura

1.32

 historically

1.30

″

1.29

AGR

1.28

Activations Density 0.001%