INDEX

Explanations

declaring variables and subroutines

This neuron activates on code-like tokens—particularly VB/VBA keywords, identifiers, punctuation, and function or subroutine definitions.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 вместе

-1.45

[{

-1.41

 después

-1.38

𓂂

-1.37

 чтобы

-1.27

 только

-1.24

mbira

-1.24

 they

-1.21

rilis

-1.19

POSITIVE LOGITS

💽

1.52

 camuf

1.34

 medarbe

1.32

 tío

1.31

赜

1.30

 novelas

1.30

 miliki

1.29

鵟

1.28

himself

1.27

themselves

1.27

Activations Density 0.221%