INDEX

Explanations

qualifiers and intensifiers

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

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

2.13

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

2.13

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

2.06

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

2.04

↵↵↵↵↵↵↵

2.04

↵↵↵↵↵↵↵↵↵↵↵↵

2.03

↵↵↵↵↵↵↵↵↵↵↵

2.03

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

2.02

↵↵↵↵↵↵↵↵↵↵

2.00

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

1.96

POSITIVE LOGITS

/=

1.43

1.41

 да

1.34

ed

1.31

 provide

1.30

opr

1.28

 stack

1.27

 deliver

1.24

 reduce

1.24

 reduced

1.24

Activations Density 0.276%