INDEX

Explanations

start of code try blocks

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

Ré

1.45

mú

1.40

пом

1.36

ח

1.35

Ré

1.33

默

1.33

 impermeable

1.33

 Mubarak

1.33

 hướng

1.32

مود

1.32

POSITIVE LOGITS

agons

1.20

abic

1.16

rations

1.06

ulloch

1.02

agan

0.98

rones

0.96

ramid

0.96

cles

0.94

bean

0.94

 jste

0.92

Activations Density 0.043%