INDEX

Explanations

being investigated, forefront of, nature of the object, agency's operations, anyone else, clear sound, data-driven system, tracing on parchment

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

ed

1.56

 function

1.55

rupt

1.55

dish

1.52

ة

1.49

dür

1.48

ла

1.42

liert

1.42

ﻭ

1.41

原地

1.41

POSITIVE LOGITS

 rehearing

1.90

UAWEI

1.70

iferous

1.67

├──

1.64

斃

1.62

Kali

1.62

Allowance

1.61

يرا

1.58

Eli

1.57

Dix

1.57

Activations Density 0.000%