INDEX

Explanations

the first step to

first step is

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

 โรง

0.46

gow

0.46

mee

0.45

 mesela

0.44

 Dhoni

0.44

 Plague

0.44

arrison

0.44

 今日

0.43

spa

0.43

 اتر

0.43

POSITIVE LOGITS

 utilization

0.57

ש

0.53

Utilization

0.52

 COMMUN

0.50

 understanding

0.50

 therein

0.50

 customarily

0.50

PORT

0.49

 PORT

0.48

ிகளை

0.48

Activations Density 0.001%