INDEX

Explanations

Larry, LR, lore

This neuron fires on capitalized name fragments beginning with “L” followed by “a,” “e,” or “o” and “r” (e.g. Lar-, Ler-, Lor-) in proper nouns.

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

 দাদা

0.58

ಲ್ಲು

0.57

 நட

0.57

nh

0.54

 들어

0.54

 ইন্ডি

0.53

லே

0.53

argaan

0.53

ຄ

0.52

 ಗೊ

0.50

POSITIVE LOGITS

azepam

0.61

 Ipsum

0.59

 நாளில்

0.56

ซ์

0.54

aneers

0.54

};

0.54

icular

0.53

 Driscoll

0.52

जों

0.52

 magnific

0.51

Activations Density 0.115%