INDEX

Explanations

immediate, quickly, instantly

The neuron consistently lights up on temporal markers—words and tokens that indicate timing, sequence, or time stamps (e.g. “use,” “fading,” “knew,” “later,” “min,” numeric minute markers).

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

 vuos

1.11

 someday

0.87

 mwaka

0.86

년간

0.86

 years

0.85

 Jahre

0.85

 vuonna

0.85

 godine

0.83

每年

0.82

 सरकार

0.82

POSITIVE LOGITS

 минут

1.52

 immediatamente

1.39

 কিছুক্ষণ

1.36

minutes

1.35

 minutes

1.30

 минуту

1.30

午前

1.28

 hurried

1.27

 फटाफट

1.27

 immediately

1.25

Activations Density 1.910%