INDEX

Explanations

size_t

This neuron detects numeric constants (literals) 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

ждается

-0.83

 کره

-0.79

Personendaten

-0.78

Jpa

-0.78

niejs

-0.75

mable

-0.73

Jpa

-0.73

笑的

-0.73

隘

-0.72

选项

-0.71

POSITIVE LOGITS

 Vorschläge

0.79

 Zuschauer

0.79

fficient

0.77

Isso

0.77

iiiiiiii

0.76

儘

0.74

 травня

0.74

 usize

0.73

 Distinction

0.73

 Residential

0.73

Activations Density 0.006%