INDEX

Explanations

This neuron picks out code identifiers—especially function, class, or variable names (often in camelCase or with underscores).

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

rières

0.77

 अल्को

0.73

 boîte

0.72

yscrapers

0.71

 ориги

0.68

 महाविद्या

0.68

ayutt

0.68

 Casualty

0.66

mateurs

0.66

rierung

0.66

POSITIVE LOGITS

0.71

<unused60>

0.63

<unused61>

0.59

<0x0D>

0.58

&&

0.57

(){

0.55

};

0.54

、

0.54

0.53

*/

0.53

Activations Density 0.030%