INDEX

Explanations

exceeding limits and comparisons

The neuron detects comparative constructions using “than,” particularly when comparing quantities or degrees (e.g., “deeper than 16 m”).

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

and

-2.27

-1.59

-1.52

but

-1.49

-1.48

-1.47

<sup>

-1.45

-1.42

-1.41

POSITIVE LOGITS

 überprü

1.79

 gewährleisten

1.72

 verurs

1.64

 beschleun

1.57

 berücksich

1.55

 Nachbarn

1.50

 calça

1.48

 Também

1.45

 varones

1.43

 défi

1.41

Activations Density 0.033%