INDEX

Explanations

describing contexts and situations

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

在

1.61

在其

1.31

 في

1.22

 रोक

1.21

在北京

1.17

in

1.16

在中

1.15

 فى

1.14

în

1.13

在国内

1.13

POSITIVE LOGITS

 accordance

2.05

 guise

1.90

 contexts

1.82

 nutshell

1.73

 vicinity

1.68

 случаях

1.67

 context

1.63

 midst

1.62

 hurry

1.61

 соответствии

1.61

Activations Density 1.179%