INDEX

Explanations

as followed by comparison or condition

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ت

2.00

т

1.91

м

1.82

tay

1.62

1.54

tion

1.52

م

1.50

es

1.49

л

1.48

ים

1.45

POSITIVE LOGITS

sembles

1.83

inine

1.61

ymmet

1.59

cribable

1.29

ignment

1.27

 וכ

1.27

 минимум

1.16

 물론

1.15

的情況

1.13

いたり

1.13

Activations Density 0.753%