INDEX

Explanations

prefer

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

 prefer

-0.99

脚注の使い方

-0.93

 prefers

-0.85

ItemBackground

-0.81

 Prefer

-0.76

 defaultstate

-0.76

ContentAsync

-0.75

 متعلقه

-0.72

prefer

-0.71

 preference

-0.70

POSITIVE LOGITS

to

0.88

ed

0.61

st

0.48

0.43

toInt

0.43

“

0.43

aires

0.42

 legacies

0.41

 Medicine

0.41

Activations Density 0.011%