INDEX

Explanations

action or process nouns

The neuron highlights substantive content words—mostly nouns and verbs that name key concepts, institutions, activities, or processes—rather than common function words.

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

سک

1.07

कर

0.98

可以看出

0.97

是大

0.96

不大

0.93

 dealings

0.92

сть

0.91

sType

0.90

হাত

0.90

行驶

0.90

POSITIVE LOGITS

 unor

1.27

 ενός

1.25

 një

1.16

1.13

eine

1.12

 sırasında

1.12

 opérés

1.11

 eines

1.10

ে

1.06

 দ্য

1.05

Activations Density 3.870%