INDEX

Explanations

listing various types of entities

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 that

-1.75

one

-1.71

 their

-1.63

any

-1.61

two

-1.53

 could

-1.52

 only

-1.49

 this

-1.47

 after

-1.46

new

-1.41

POSITIVE LOGITS

 både

1.54

 zarówno

1.40

both

1.37

 diverses

1.34

Including

1.34

 verschieden

1.33

numer

1.29

rungen

1.27

ranging

1.27

cluding

1.27

Activations Density 0.085%

No Known Activations