INDEX

Explanations

Single letters

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

 including

-0.54

 with

-0.53

 Including

-0.50

 himo

-0.50

 CreateTagHelper

-0.48

AsUp

-0.45

 Italij

-0.45

Including

-0.45

 ervan

-0.43

POSITIVE LOGITS

etheless

0.64

 surla

0.61

}")

0.54

COUVER

0.54

//

0.54

$.

0.52

UREAU

0.52

Portail

0.52

hamshire

0.52

};*/

0.52

Activations Density 0.004%