INDEX

Explanations

Kiev, Walker, Uemura

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

çĴ§

-0.23

 SQLite

-0.23

æ¡ħ

-0.23

 initials

-0.23

emacs

-0.22

IGNORE

-0.22

edList

-0.22

 nursery

-0.22

 misdemean

-0.22

 pancakes

-0.22

POSITIVE LOGITS

 Incre

0.28

æ²»çĸĹ

0.27

ients

0.26

_bases

0.25

seo

0.25

.builders

0.24

ç§ĳåŃ¦å®¶

0.24

glob

0.24

.getParam

0.23

éĺ³æĢ§

0.23

Activations Density 0.004%