INDEX

Explanations

chamber

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asser

-0.32

éĢ¾

-0.27

 Anchor

-0.26

å¹³è¡Į

-0.26

 Rendering

-0.26

ç¯Ļ

-0.26

Ã©rt

-0.26

ponsors

-0.25

è¢Ń

-0.25

 Renderer

-0.25

POSITIVE LOGITS

ing

0.42

ed

0.38

edBy

0.29

å±ŀå®ŀ

0.28

itious

0.28

ming

0.27

caps

0.25

ble

0.25

cies

0.25

èµ·äºĨ

0.25

Activations Density 0.392%