INDEX

Explanations

to kill

AI platform designs

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ної

0.45

 رج

0.45

 ചു

0.44

၊

0.44

igheder

0.44

 особли

0.44

 يجعل

0.43

来

0.43

があれば

0.42

machen

0.42

POSITIVE LOGITS

 replicate

0.43

 stimulate

0.42

 mathematicians

0.39

RIBUTE

0.39

 analyze

0.39

 calculate

0.38

 bakers

0.38

 unconditionally

0.38

 bardziej

0.38

 spreadsheets

0.38

Activations Density 0.000%