INDEX

Explanations

discrepancies or differences

This neuron detects warnings about slight variations or deviations in a product’s displayed appearance—especially color differences across screens or monitors.

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

着一个

-1.46

 fermés

-1.40

 senador

-1.40

 juiz

-1.24

ੂੰ

-1.20

 associées

-1.19

釭

-1.17

шни

-1.16

 animés

-1.15

viu

-1.13

POSITIVE LOGITS

our

1.59

 their

1.57

 there

1.43

 other

1.39

 small

1.34

 these

1.31

 people

1.28

īgi

1.27

 those

1.27

 кровать

1.23

Activations Density 0.014%