INDEX

Explanations

document structure

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 strugg

-0.24

æ¸¸æĪıæĵįä½ľ

-0.21

 ofApp

-0.21

nyder

-0.21

 ADVISED

-0.21

åĹ«

-0.21

rvine

-0.21

incinn

-0.21

çļĦæ°Ķæģ¯

-0.21

 Ø³Ø¨ØŃØ§ÙĨÙĩ

-0.20

POSITIVE LOGITS

.Does

0.23

 ÑĥÐ²ÐµÐ»Ð¸ÑĩÐ¸

0.22

doi

0.20

å·¦

0.20

.Today

0.20

 Bold

0.19

:left

0.19

 legend

0.19

éķĲ

0.19

.StatusOK

0.19

Activations Density 0.089%

No Known Activations