INDEX

Explanations

Palmer, Carr, Richardson names

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 unſ

-2.64

The

-2.59

鶿

-2.58

꯶

-2.52

퐹

-2.47

裥

-2.41

 accompanying

-2.38

 unveiling

-2.36

ㆆ

-2.34

 enchanting

-2.33

POSITIVE LOGITS

2.91

↵

2.53

2.33

↵↵

2.30

～

2.22

 étant

2.17

銼

2.17

雠

2.14

</em>

2.13

难怪

2.11

Activations Density 0.005%