INDEX

Explanations

code and data analysis

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

 Manus

-0.26

ÑĩÑĥÐ²

-0.25

åı¦ä¸Ģè¾¹

-0.25

éŃħåĬĽ

-0.24

contexts

-0.24

_yaml

-0.24

edList

-0.24

 uyar

-0.24

preset

-0.23

Interfaces

-0.23

POSITIVE LOGITS

 data

0.64

æķ°æį®

0.60

data

0.51

ãĥĩãĥ¼ãĤ¿

0.49

æķ¸æĵļ

0.47

æķ°æį®åĪĨæŀĲ

0.45

 Ð´Ð°Ð½Ð½ÑĭÐµ

0.44

 dá»¯

0.43

	data

0.43

.data

0.42

Activations Density 0.019%