INDEX

Explanations

The main thing this neuron does is detect numeric tokens (numbers, units, and measurement expressions).

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

ියාව

0.66

magazine

0.65

ʿ

0.57

 অধিদপ্তর

0.54

 عظ

0.53

 Penit

0.53

homepage

0.52

Magazine

0.52

telek

0.51

絢

0.51

POSITIVE LOGITS

/+

0.73

0.72

non

0.65

 kurulum

0.64

 freestanding

0.63

 üzeri

0.63

 arası

0.63

($

0.62

 opciones

0.61

 пользователь

0.61

Activations Density 0.334%