INDEX

Explanations

can browse or select

can find details

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 liefern

0.54

 morrer

0.50

 producir

0.48

 raping

0.47

 entregar

0.45

 produisent

0.43

 funcionar

0.42

 bieten

0.42

 bombing

0.42

 bombings

0.42

POSITIVE LOGITS

 browse

1.02

 check

1.00

 explore

0.89

 download

0.88

 consult

0.88

 peruse

0.85

 choose

0.81

 checkout

0.81

 consider

0.79

browse

0.77

Activations Density 0.752%