INDEX

Explanations

identify, track, or provide information

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

это

1.71

というと

1.67

 это

1.66

this

1.65

 Tatsache

1.63

 λοι

1.61

我们就

1.58

这么

1.57

 именно

1.55

的话

1.54

POSITIVE LOGITS

 revitalize

1.23

 bolster

1.23

 foreclose

1.18

 rebounded

1.15

 tonal

1.13

ubine

1.12

 fetal

1.12

 minimax

1.11

␤

1.09

enchen

1.09

Activations Density 0.261%