INDEX

Explanations

draft pick or selection

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

Bib

-0.24

åĲĿ

-0.23

é»ĦçŁ³

-0.23

 considerable

-0.23

 glowing

-0.22

çĭĲ

-0.22

Violation

-0.21

UpperCase

-0.21

nces

-0.21

esome

-0.21

POSITIVE LOGITS

xr

0.27

 inaugural

0.25

åħ¥èģĮ

0.25

 Optional

0.23

åħ¬å¼Ģ

0.23

 Naming

0.23

 optional

0.23

è¯ģçĽĳä¼ļ

0.23

 Poll

0.23

 Pall

0.22

Activations Density 0.006%