INDEX

Explanations

rights reserved or likelihood of

The neuron spikes on punctuation tokens—especially the colon character “:” (as seen in time‐format strings and timestamps).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

routines

-0.80

رام

-0.78

 Eugen

-0.69

 concerne

-0.68

Sony

-0.68

Leon

-0.67

Euro

-0.66

﹎

-0.66

lucene

-0.66

dw

-0.65

POSITIVE LOGITS

 slash

0.76

roits

0.75

änden

0.75

work

0.71

 smoothie

0.70

 gondola

0.70

Basta

0.69

 acceptance

0.69

 emperors

0.68

 SwitchCompat

0.68

Activations Density 0.067%