INDEX

Explanations

code and documentation titles

The neuron strongly responds to HTML (and XHTML) document title elements—i.e. the `<title>` tag and its enclosed text.

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

the

-1.70

 gorra

-1.43

hâ

-1.40

鹩

-1.40

 diverses

-1.36

 lourdes

-1.36

 deine

-1.35

を読む

-1.33

 personalización

-1.33

 thre

-1.30

POSITIVE LOGITS

 after

1.73

 with

1.73

 will

1.61

or

1.52

for

1.49

in

1.47

 although

1.44

 große

1.43

but

1.41

 would

1.40

Activations Density 0.387%