INDEX

Explanations

compile, instructions, registers, bytecode

The neuron fires on technical software‐development terms referring to compiled code artifacts and related operations (e.g. “class,” “file,” “decompile,” “disassemble,” “reverse‐engineer”).

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

bergement

0.63

 প্রশাসনিক

0.63

ipheral

0.59

 físicas

0.59

izquierda

0.56

 trá

0.56

PHYS

0.55

 kancel

0.55

 했습니다

0.55

كامل

0.54

POSITIVE LOGITS

JVM

0.78

 Java

0.73

JVM

0.70

 bytecode

0.68

 JAVA

0.65

 java

0.63

 synthetic

0.62

 registers

0.61

Java

0.61

 monit

0.59

Activations Density 0.015%