INDEX

Explanations

journal abbreviations and ". "

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

æĳĨ

-0.25

iff

-0.20

éĤ¦

-0.20

 effort

-0.19

olls

-0.19

mente

-0.19

omorphic

-0.19

ä¸İåħ¶

-0.18

é¦

-0.18

åĲĪå½±

-0.18

POSITIVE LOGITS

 newcom

0.24

quo

0.20

æĢİä¹Īçľĭ

0.20

è¯ĺ

0.19

orque

0.19

æĹ¥æĬ¥

0.19

_MESSAGES

0.18

 Minute

0.18

toolbox

0.18

ogl

0.17

Activations Density 0.029%