INDEX

Explanations

codes or variable assignments

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

ree

-0.07

 faithfully

-0.07

è¢ħè¢ħ

-0.07

rees

-0.07

.isOn

-0.07

prototype

-0.06

 accurate

-0.06

ä¹Łä¸įå¤ļ

-0.06

ique

-0.06

 separately

-0.06

POSITIVE LOGITS

sz

0.08

icode

0.08

unei

0.07

codes

0.07

Codes

0.07

ICODE

0.07

amenta

0.07

è·¯éĢı

0.07

ä¸įæľį

0.07

åĨħå®¹æľī

0.07

Activations Density 0.013%