INDEX

Explanations

java, sql, crypto, apache, dspace, quarkus, eclipse

This neuron activates on the top‐level package identifiers (like “org,” “javax,” “java,” “com,” etc.) in Java import or package statements.

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

 וכן

-1.29

 latter

-1.02

 diğer

-0.96

};

-0.96

if

-0.93

อื่น

-0.91

or

-0.89

逃避

-0.89

 salonu

-0.89

 mall

-0.88

POSITIVE LOGITS

</u>

1.14

 champignons

1.09

から

1.08

隄

1.05

 fatis

1.04

 dacia

1.02

 paff

1.02

шёл

1.01

 nobile

0.99

</b>

0.98

Activations Density 0.005%