INDEX

Explanations

This neuron detects numerical expressions—especially floating‐point or large numeric values (e.g. statistical counts) 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

No Comments

Negative Logits

 Billing

0.95

 Contact

0.90

 Mesmo

0.89

 Engaging

0.85

 Preparing

0.83

 CONTACT

0.83

 Consultado

0.82

 Pending

0.82

 collaborators

0.81

 Calculates

0.81

POSITIVE LOGITS

 subjekt

0.82

గ్రహ

0.81

 termen

0.70

dû

0.70

Energie

0.69

the

0.68

 statut

0.68

perme

0.68

les

0.68

Produkt

0.68

Activations Density 0.000%