INDEX

Explanations

government regulations and publications

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

-1.73

we

-1.47

,’

-1.44

↵↵

-1.38

out

-1.36

 aislados

-1.35

 funciona

-1.34

.’

-1.30

khi

-1.25

 pelos

-1.23

POSITIVE LOGITS

 armon

1.38

洐

1.35

 usarlo

1.34

 regata

1.34

 maniere

1.30

 Issus

1.29

quettes

1.27

眈

1.27

 boute

1.26

 pantalon

1.24

Activations Density 0.028%