INDEX

Explanations

The neuron fires on code‐ or markup‐style structural tokens—e.g. arrow operators (->, =>), namespace declarations, punctuation like | and &, and similar symbols.

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

则

0.70

 thì

0.66

貥

0.63

则是

0.61

="";

0.61

 however

0.61

則

0.61

="(

0.60

 Thì

0.59

#"

0.59

POSITIVE LOGITS

 setattr

0.78

»).

0.64

 '').

0.64

 '');

0.62

»),

0.62

{})

0.61

）、

0.60

»)

0.60

 impresa

0.59

 bract

0.58

Activations Density 0.000%