INDEX

Explanations

1

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

äº²çľ¼

-0.29

ify

-0.28

è¯ºè´Ŀå°Ķ

-0.28

quire

-0.27

æĦ§

-0.27

quiring

-0.27

tÃł

-0.26

.getHeader

-0.25

æµĵæµĵçļĦ

-0.25

 Nobel

-0.24

POSITIVE LOGITS

-than

0.31

-com

0.28

åĲĪè§Ħ

0.27

å¢¨è¥¿åĵ¥

0.25

-turn

0.25

ÐµÐºÑĤÐ¸Ð²

0.25

 ÑĤÑĢÑĥÐ´

0.25

 Clean

0.25

fore

0.24

_COMP

0.24

Activations Density 0.005%