INDEX

Explanations

resource acquisition

The neuron activates on decimal numbers (floating‐point values) in the text.

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

</strong>

0.73

 cinquième

0.73

üm

0.72

 lançamento

0.71

و

0.70

</h1>

0.70

 kwietnia

0.70

<0x80>

0.69

 kwiet

0.69

 TRIAL

0.69

POSITIVE LOGITS

 demeanor

0.74

 endlessly

0.73

провод

0.70

 brazen

0.68

ыска

0.67

పంచ

0.67

 whoever

0.66

 beset

0.66

 unorthodox

0.66

εις

0.66

Activations Density 0.000%