INDEX

Explanations

young person

The neuron fires on numeric age mentions (e.g. “35-year-old,” “9-Year-Old,” “old” in “year-old” constructions).

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

 repertoire

0.86

 photosynthesis

0.84

 triplicate

0.82

 wonderland

0.79

 readable

0.77

 masterpiece

0.77

 storied

0.76

 masala

0.76

 sublime

0.75

 homeostasis

0.75

POSITIVE LOGITS

 fiatal

1.02

 women

0.98

 مسلمان

0.97

 personer

0.96

 young

0.95

women

0.95

 teenagers

0.95

 jeunes

0.94

 clergyman

0.94

 teenager

0.94

Activations Density 0.203%