INDEX

Explanations

numbers in code contexts

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

 δύο

-1.53

缦

-1.48

 sogen

-1.48

粁

-1.43

weren

-1.42

 geringen

-1.41

畯

-1.38

 technologically

-1.36

蹀

-1.36

ujeme

-1.34

POSITIVE LOGITS

1.39

1.32

1.29

of

1.27

1.24

1.23

1.13

なくな

1.13

).

1.11

сить

1.09

Activations Density 0.061%