INDEX

Explanations

code related to implementations or specific classes

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

ersch

-0.07

iliation

-0.07

ä¸ĩè¾¾

-0.07

Ã¼p

-0.07

 Twice

-0.07

zas

-0.07

zor

-0.07

iges

-0.06

uken

-0.06

igest

-0.06

POSITIVE LOGITS

æļĤè¡Į

0.08

 Donn

0.08

Ð¿ÑĢÑĥ

0.07

alize

0.06

.emit

0.06

æľĭ

0.06

åį³åı¯æŁ¥çľĭ

0.06

ContentView

0.06

ç¦»çº¿

0.06

alized

0.06

Activations Density 0.892%