INDEX

Explanations

`INSERT INTO` or `CREATE INDEX`

The neuron is triggered by HTML tag and attribute tokens (e.g. “a”, “i”, “aria‐hidden”).

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

 अल्को

0.66

 अभिकारक

0.64

स्तू

0.62

<unused81>

0.60

Além

0.60

 सूत्रकृ

0.59

<unused47>

0.59

 व्हाउचर

0.58

ÈRE

0.57

 लेणी

0.57

POSITIVE LOGITS

0.63

="

0.59

0.58

 width

0.58

 className

0.56

 charset

0.56

0.55

0.54

css

0.53

Activations Density 0.016%