INDEX

Explanations

early advantage

startup and starting things

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 began

1.59

 begun

1.47

 begy

1.30

 begins

1.27

 BEGIN

1.23

 begyn

1.23

 beginning

1.21

 begin

1.20

 zač

1.15

Begin

1.13

POSITIVE LOGITS

 Starter

2.50

 starter

2.47

 Start

2.42

 startup

2.42

 Startup

2.41

Startup

2.33

starter

2.31

Starter

2.30

 START

2.28

 стар

2.28

Activations Density 0.142%