INDEX

Explanations

powers of terms in parentheses

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

anj

-0.08

iband

-0.08

åŃľåŃľ

-0.07

_unused

-0.07

etine

-0.07

udit

-0.07

attery

-0.07

ëĭĪëĭ¤

-0.07

åĿļæĮģä¸įæĩĪ

-0.07

Ð¾ÑĤÐ¾Ð²

-0.07

POSITIVE LOGITS

picker

0.07

attra

0.07

TargetException

0.07

 Picker

0.07

åı¯éĢīä¸Ń

0.06

å¾Ļ

0.06

è¢ĸ

0.06

 opener

0.06

çļĦæīĭèĩĤ

0.06

åĶ§

0.06

Activations Density 0.002%