INDEX

Explanations

start of phrase or list item

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 attorneys

-0.81

кистан

-0.79

入院

-0.79

 Electron

-0.75

 pemain

-0.75

лец

-0.75

 chargée

-0.73

лекса

-0.73

tifik

-0.72

expectations

-0.71

POSITIVE LOGITS

VET

0.85

 industry

0.84

 qualifications

0.79

떡

0.78

 trainer

0.78

VET

0.78

니메

0.77

 copolymer

0.77

 qualification

0.75

straining

0.75

Activations Density 0.007%