INDEX

Explanations

methods and techniques

The neuron detects mentions of methods or techniques—words like “used,” “employed,” “techniques,” and similar—especially in technical or academic 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

boyfriend

-1.54

优秀的

-1.45

 ayudó

-1.41

karena

-1.40

pulseira

-1.40

 llegó

-1.40

biker

-1.36

錚

-1.36

 entró

-1.36

جهة

-1.36

POSITIVE LOGITS

to

2.23

 techniques

1.88

 methods

1.77

–

1.68

.”

1.60

蓠

1.59

 measures

1.58

快要

1.57

ју

1.55

その

1.54

Activations Density 0.397%