INDEX

Explanations

predicting descriptive words after conjunctions

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 axiomatic

1.72

 vacuo

1.64

ittal

1.63

品の

1.58

;}

1.58

식회사

1.57

 полном

1.57

лес

1.55

*}

1.54

"};

1.50

POSITIVE LOGITS

 many

1.77

 replacements

1.56

 replacement

1.46

 Vari

1.44

many

1.43

 prospects

1.40

 possibility

1.40

pport

1.39

 varied

1.39

 variety

1.38

Activations Density 0.016%