INDEX

Explanations

numbers separated by punctuation

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 rootNode

-0.25

è¹ī

-0.24

BAT

-0.23

ä»½

-0.22

å²Ĺ

-0.22

PerPixel

-0.22

 equiv

-0.21

èĬĥ

-0.21

FF

-0.21

ä¸Ģä¸ªå¤ļ

-0.21

POSITIVE LOGITS

æļĤæĹ¶

0.25

.hamcrest

0.23

aviest

0.21

thing

0.21

.communication

0.21

scn

0.21

.uc

0.21

ayan

0.21

decimal

0.20

æ¸©æŁĶ

0.20

Activations Density 0.116%

No Known Activations