INDEX

Explanations

analysis of predictors

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

Comentários

-0.94

 dose

-0.88

Elaboración

-0.87

sm

-0.85

锱

-0.83

+--

-0.82

Preparazione

-0.82

 ernst

-0.80

Caracteristici

-0.80

Saiba

-0.79

POSITIVE LOGITS

 amitié

0.90

vânt

0.89

自分が

0.85

sunter

0.85

 viață

0.84

avantage

0.83

0.81

 лечение

0.81

 céramique

0.79

Activations Density 0.000%