INDEX

Explanations

of

The neuron consistently lights up on quantifying terms and expressions—like “one,” “half,” “each,” “some,” and “cases”—used to enumerate or measure elements of a narrative.

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

到一个

0.77

 reveal

0.65

ंनी

0.63

NewGuid

0.63

Jährige

0.62

 efficiently

0.62

 Polsce

0.61

ברים

0.61

さんが

0.60

CriteriaUtils

0.59

POSITIVE LOGITS

 của

4.72

of

4.59

 ofthe

4.59

 của

4.31

ของ

4.25

 של

3.73

 της

3.45

 των

3.38

ຂອງ

3.38

ofthe

3.23

Activations Density 3.866%