INDEX

Explanations

information requests and records

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

-1.43

-1.42

−

-1.37

鞏

-1.32

 guerr

-1.30

土産

-1.21

There

-1.21

</h2>

-1.20

 péché

-1.20

on

-1.18

POSITIVE LOGITS

 crescere

1.56

 garanzia

1.49

 needed

1.47

氽

1.46

 musíte

1.45

 capitolo

1.44

姤

1.41

歺

1.40

 cammino

1.40

 dedicato

1.40

Activations Density 0.044%