INDEX

Explanations

displays patterns or showed that

The neuron fires on verbs that report or present findings—e.g. “show,” “display,” “interpret,” “reveal”—typically used to describe results in scientific or analytical contexts.

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

of

-1.99

-1.89

姽

-1.72

姈

-1.70

嫒

-1.69

ми

-1.67

низу

-1.66

 cinque

-1.66

إذا

-1.65

ſſen

-1.60

POSITIVE LOGITS

</h3>

1.66

 their

1.64

1.63

灕

1.59

 blames

1.55

</h4>

1.55

ཫ

1.50

 desnuda

1.49

Carsten

1.48

ся

1.45

Activations Density 0.005%