INDEX

Explanations

single or one

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

 সংখ

-0.07

 thinner

-0.07

 THAT

-0.06

 muffin

-0.06

あくまで

-0.06

 তালিক

-0.06

 ONLY

-0.06

ONLY

-0.06

নির্দিষ্ট

-0.06

 begged

-0.06

POSITIVE LOGITS

 চতু

0.06

धान

0.06

ginx

0.06

 denominado

0.06

<0xB4>

0.05

调用

0.05

 spacious

0.05

 errone

0.05

 эффект

0.05

 इन्हें

0.05

Activations Density 0.009%