INDEX

Explanations

error handling and clear explanations

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

 maintain

1.25

 abound

1.16

 assume

1.09

asso

1.07

ojn

1.04

 सध्या

1.03

甚至

1.02

τικού

1.02

하거나

1.01

ನ್ಯ

1.01

POSITIVE LOGITS

finally

1.89

終於

1.78

终于

1.67

 endlich

1.61

 наконец

1.60

конец

1.49

 finalmente

1.47

 نهايه

1.43

 finally

1.41

ENDPOINT

1.40

Activations Density 0.290%