INDEX

Explanations

tables with features

The neuron primarily activates on numeric tokens and numeric data (e.g., numbers, measurements, and decimal values).

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

 fürs

0.64

 również

0.60

);

0.57

感谢

0.57

也是

0.56

西北

0.55

すぎる

0.55

邮件

0.55

captcha

0.55

Although

0.54

POSITIVE LOGITS

(%)

0.77

 иной

0.59

0.56

$/

0.56



0.56

 $\%$

0.55

 പ്രതീ

0.55

_|

0.55

 יכול

0.55

 }}{\

0.54

Activations Density 0.020%