INDEX

Explanations

research

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

出版年

-0.71

 invokingState

-0.65

adpleegd

-0.63

windowFixed

-0.61

 utafitiHapana

-0.56

Beat

-0.56

 חיצוניים

-0.55

UnusedPrivate

-0.54

Bole

-0.53

beat

-0.52

POSITIVE LOGITS

 sauvage

0.51

WarningLevel

0.50

rogels

0.46

Inherited

0.45

 Audiodateien

0.45

 OnInit

0.45

uitton

0.44

 Lokal

0.44

 layui

0.44

Virgo

0.43

Activations Density 0.013%