INDEX

Explanations

asking for specific advice

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

 crippling

1.37

 notorious

1.34

 fatal

1.33

 pernicious

1.30

 infamous

1.26

 debilitating

1.25

 ineffective

1.24

.}}

1.23

excellent

1.21

 violently

1.20

POSITIVE LOGITS

Maybe

1.72

 Maybe

1.68

 poderia

1.67

 pourrait

1.64

 آینده

1.64

 yapmak

1.62

 futura

1.60

 haber

1.59

的可能性

1.58

 podría

1.57

Activations Density 0.020%