INDEX

Explanations

handle exceptions or processes

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

æĸ¹æ³ķè¿Ľè¡Į

-0.07

mot

-0.07

Ð¿ÐµÐ½Ð¸Ñı

-0.07

 borderTop

-0.06

 borderBottom

-0.06

 presta

-0.06

 trÃºc

-0.06

boy

-0.06

gaard

-0.06

 anytime

-0.06

POSITIVE LOGITS

 instead

0.06

amm

0.05

OR

0.05

 zami

0.05

instead

0.05

 alma

0.05

 Instead

0.05

åı¦

0.05

çģµéŃĤ

0.05

ä»£æĽ¿

0.05

Activations Density 0.013%