INDEX

Explanations

a) or a

The neuron is keyed to detect numeric list markers (e.g. question/item numbers and blanks) in the text.

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

sku

0.63

ziemy

0.63

larımız

0.58

 oraz

0.57

 sowie

0.56

 queremos

0.56

 prad

0.56

 esperamos

0.56

 aspires

0.55

stven

0.55

POSITIVE LOGITS

𝕋

0.64



0.62

selective

0.61

А

0.60

blogs

0.59

εν

0.59

 ಆನಿ

0.59

①

0.59

キング

0.58

𝘼

0.58

Activations Density 0.004%