INDEX

Explanations

semicolon followed by it

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

ete

2.78

欖

2.37

oshop

2.35

eted

2.34

兩

2.30

ureate

2.29

idoscope

2.21

त्ते

2.18

ett

2.18

शुदा

2.14

POSITIVE LOGITS

4.45

т

3.41

sion

3.40

sens

3.39

3.10

sap

3.07

san

3.04

sons

2.86

sas

2.80

sü

2.67

Activations Density 0.111%