INDEX

Explanations

dates and amounts

The neuron fires strongly on mentions of famous work titles (e.g. “Sunshine” the song or “Harry Potter” the book series).

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

 मच

0.55

ducible

0.54

edor

0.53

του

0.51

த்தக

0.51

াত্র

0.51

 وزير

0.51

**/

0.50

 isActive

0.50

rman

0.50

POSITIVE LOGITS

∉

0.62

0.59

聽

0.57

 Keene

0.56

Falling

0.54

Scholar

0.52

Squirrel

0.52

致

0.52

ână

0.51

ᴥ

0.51

Activations Density 0.030%