INDEX

Explanations

either express or implied

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

 gauge

-0.78

翳

-0.77

 größe

-0.77

ActionBar

-0.75

 избежать

-0.75

Allemagne

-0.73

ול

-0.73

('')

-0.72

ό

-0.71

 tight

-0.70

POSITIVE LOGITS

ின்ற

0.83

 volta

0.83

]]

0.82

 causa

0.79

Lecture

0.78

 poder

0.74

𖥻

0.74

 pneumoniae

0.73

Rgb

0.73

ÉR

0.73

Activations Density 0.071%