INDEX

Explanations

"en línea recta" or "basis"

The neuron activates on legal disclaimer language—especially “as is” service clauses, warranty disclaimers, liability releases, and indemnification phrasing.

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

ফুল্ল

0.59

ilites

0.57

升高

0.56

iculata

0.55

ဲ့

0.54

र्सेज

0.54

రుగులు

0.52

 мили

0.51

ሁ

0.51

 आतंकी

0.51

POSITIVE LOGITS

 basis

1.48

basis

1.38

 Basis

1.23

Basis

1.16

 BASIS

1.09

 Grundlage

1.01

BAS

0.85

 ভিত্তিতে

0.84

black

0.83

 black

0.82

Activations Density 0.145%