INDEX

Explanations

code includes

The neuron is primarily triggered by numeric literals (numbers) in code.

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

después

-1.34

patrón

-1.30

árbol

-1.30

según

-1.25

Ꮸ

-1.23

Ꮀ

-1.22

 colchón

-1.21

にほんブログ村

-1.18

のおかげ

-1.17

 bordado

-1.16

POSITIVE LOGITS

—

1.20

↵↵

0.96

 also

0.95

 often

0.90

0.88

此外

0.87

 Übersetzungen

0.87

Activations Density 0.015%