INDEX

Explanations

useless or unused things

The neuron fires on hedging or advisory phrasing—especially modal verbs (“can,” “could”) and strong cautionary words (“extremely,” “overwhelming,” “costly”) that flag warnings or recommendations.

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

 الكامل

0.71

 Allerdings

0.71

満足

0.66

 موفق

0.66

 superiore

0.64

更に

0.63

 کامل

0.63

 خفی

0.63

WithOptions

0.63

 başarılı

0.63

POSITIVE LOGITS

 landfills

1.02

 repurposed

1.02

 wasteful

1.00

 useless

0.99

 disposable

0.99

 unused

0.98

 reused

0.98

 inutil

0.97

 utensils

0.96

 unusable

0.96

Activations Density 0.136%