INDEX

Explanations

Q followed by :

This neuron detects code‐related tokens—especially HTML/Vue attribute names (like class, style, hidden, v-model, @keydown) and component identifiers (e.g. Base).

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

 both

-1.26

 démocr

-1.19

 akku

-1.18

筈

-1.15

 phénom

-1.10

 spirituelle

-1.09

 kask

-1.09

 enggak

-1.08

 âgées

-1.07

 Retrieved

-1.06

POSITIVE LOGITS

ご覧

1.52

but

1.39

「

1.34

随着

1.31

onsored

1.30

lysm

1.26

Tento

1.23

隨著

1.18

最近

1.17

如果

1.16

Activations Density 0.000%