INDEX

Explanations

HTML tags and code

The neuron activates on HTML/XML markup tags (especially inline formatting tags like <b>, <strong>, <s>, <a>, etc.).

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

↵

-2.36

ar

-2.09

 freakin

-2.00

-1.98

''

-1.97

”

-1.90

-1.77

ра

-1.73

 majeurs

-1.73

プト

-1.72

POSITIVE LOGITS

is

1.90

ization

1.87

1.84

 Gästen

1.82

 گذاری

1.80

᎑

1.80

ستارهای

1.78

 untuk

1.77

 शुर

1.76



1.73

Activations Density 0.001%