INDEX

Explanations

results and theorems

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

ç«ĻåĨħæĲľç´¢

-0.08

åħļæĹĹ

-0.08

æĸ°åĬłåĿ¡

-0.08

etil

-0.08

 sitÃ¤

-0.07

ë¡

-0.07

olls

-0.07

ÙĥØ§

-0.07

NewLabel

-0.07

way

-0.07

POSITIVE LOGITS

åħ³äºİ

0.08

 regarding

0.08

 about

0.08

 bahwa

0.07

theorem

0.07

éĹľæĸ¼

0.07

 theorem

0.07

Ø§Ø«

0.06

è§Ħå¾ĭçļĦ

0.06

à¹Ģà¸ģà¸µà¹Īà¸¢à¸§à¸ģà¸±à¸ļ

0.06

Activations Density 0.035%