INDEX

Explanations

quantifiers and adjectives

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

variation

-0.23

èĦĶ

-0.21

 recruiters

-0.21

ipel

-0.21

 factorial

-0.21

recommend

-0.20

çī¹åĪ«å£°æĺİ

-0.20

çļĦæ¯įäº²

-0.20

\Tests

-0.19

 æĸ¹æ³ķ

-0.19

POSITIVE LOGITS

ifiable

0.24

uess

0.23

 hopefully

0.22

æĪĸå¤ļ

0.22

xb

0.22

_THAT

0.22

ylan

0.21

xC

0.21

ings

0.21

éĹ´æİ¥

0.21

Activations Density 0.567%