INDEX

Explanations

assigning blame, guilt, or accountability

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

 văn

0.93

immersive

0.91

 festge

0.91

 diesem

0.90

 přes

0.89

zg

0.88

탑

0.87

 সম্মত

0.86

 jurnal

0.86

 dicas

0.85

POSITIVE LOGITS

 blaming

1.81

 blame

1.77

 blames

1.73

 blamed

1.52

 culprits

1.51

 culpa

1.50

 culprit

1.47

 condemnation

1.43

탓

1.40

 helplessness

1.37

Activations Density 0.361%