INDEX

Explanations

line, lines

This neuron responds to the token “line” (including instances like “line,” “Line,” or as part of words such as “online” or “Lineage”).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ომ

0.69

 glas

0.66

Mej

0.61

 அளித்த

0.61

 IANS

0.61

EME

0.60

oum

0.58

లే

0.57

 Mays

0.57

rito

0.57

POSITIVE LOGITS

 Line

1.70

 line

1.59

Line

1.55

 LINE

1.53

 Lines

1.44

 lines

1.43

 LINES

1.38

LINE

1.38

 linea

1.37

 линия

1.37

Activations Density 0.149%