INDEX

Explanations

as to

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

çļĦå·¨å¤§

-0.10

éĥľ

-0.09

çļĦè§ĤçĤ¹

-0.09

ä¸Ģåıªæīĭ

-0.09

ç´¯ç´¯

-0.09

TÄ°

-0.08

æī¿è®¤

-0.08

åİŁè°ħ

-0.08

ç´ĭ

-0.08

çº¹

-0.08

POSITIVE LOGITS

æĢİæł·

0.14

how

0.13

å¦Ĥä½ķ

0.13

æĺ¯å¦Ĥä½ķ

0.12

æĢİä¹Īæł·

0.12

æĺ¯æĢİæł·

0.11

how

0.11

 whether

0.11

æĺ¯æĢİä¹Ī

0.10

æĺ¯åĲ¦

0.10

Activations Density 0.140%