INDEX

Explanations

whether something is true

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

 consum

-0.10

 Ranger

-0.10

 Bracket

-0.10

person

-0.10

iger

-0.10

aud

-0.10

 Cherry

-0.10

 reasonable

-0.09

 curb

-0.09

 unrealistic

-0.09

POSITIVE LOGITS

 sure

0.34

 certainty

0.26

sure

0.25

 Sure

0.23

Sure

0.19

 assures

0.18

 assured

0.17

ç¡®å®ļ

0.17

 assure

0.16

 guarantee

0.16

Activations Density 0.254%