INDEX

Explanations

change negative patterns

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

 deft

1.12

 vehement

1.10

 lackluster

1.08

 rejection

1.07

 revelations

1.06

 explos

1.06

 devolution

1.05

 geochemical

1.05

 geochemistry

1.03

娇

1.03

POSITIVE LOGITS

udor

0.96

assurer

0.84

 combatir

0.80

uur

0.79

 personnes

0.77

incent

0.77

erre

0.76

 residuos

0.76

lasma

0.76

nus

0.76

Activations Density 0.151%