INDEX

Explanations

programming class member access

The neuron fires on library or framework identifiers written in CamelCase (e.g. class/utility/manager/loader 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

Negative Logits

 with

-1.91

all

-1.91

and

-1.82

 each

-1.80

 every

-1.73

on

-1.70

in

-1.63

ına

-1.46

any

-1.42

Each

-1.38

POSITIVE LOGITS

 EIGHT

1.59

<bos>

1.44

 SEVEN

1.43

 cependant

1.43

 habituellement

1.37

 généralement

1.35

 ACTUALLY

1.35

 Russel

1.31

 rassemb

1.30

有些不

1.30

Activations Density 0.013%