INDEX

Explanations

Hart

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

 Hart

-1.14

Hart

-0.94

 createState

-0.56

 HART

-0.56

 MonoBehaviour

-0.55

ār

-0.53

ReusableCell

-0.53

WebControls

-0.52

StoreMessageInfo

-0.52

CopyWith

-0.51

POSITIVE LOGITS

 Silurian

0.60

 nucléaire

0.56

 Postmaster

0.55

anyahu

0.55

ifrance

0.55

 الحره

0.54

 Devonian

0.53

 didik

0.53

 Esau

0.53

lovakia

0.52

Activations Density 0.058%