INDEX

Explanations

than or compared to

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 jiné

-1.08

 homicidio

-0.91

 meteen

-0.85

 столь

-0.85

 kín

-0.85

 μεγά

-0.84

 další

-0.84

SaveChangesAsync

-0.83

 ofrecemos

-0.83

 andere

-0.81

POSITIVE LOGITS

 than

3.45

than

1.48

now

1.30

 when

1.26

жели

1.18

Than

1.16

 compared

1.09

 because

1.06

cionada

1.06

 теперь

1.04

Activations Density 0.034%