INDEX

Explanations

corner points and sides

The neuron fires strongly on quantitative specifications—numbers, measurements, and units (e.g. dimensions, quantities, dosages) in technical or instructional 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

 Ormond

0.57

 स्था

0.52

🏺

0.52

 submer

0.52

Colon

0.52

Radial

0.52

ஶ

0.50

 Principle

0.50

 pourcentage

0.49

 globules

0.49

POSITIVE LOGITS

 sides

1.45

 corners

1.27

 Sides

1.24

sides

1.23

 lados

1.20

Sides

1.18

 vertices

1.16

 diagonals

1.16

 edges

1.13

 hexagon

1.12

Activations Density 0.155%