INDEX

Explanations

location within or near bodies of water

The neuron detects geographic and locational descriptors—words indicating place names, borders, directional positions, and related context.

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

原始内容存档于

-2.11

 then

-1.73

 using

-1.65

 when

-1.59

celotti

-1.55

Dış

-1.48

 nové

-1.44

 பய

-1.41

hidrat

-1.41

by

-1.41

POSITIVE LOGITS

1.49

скур

1.25

rous

1.17

of

1.17

ୀ

1.15

 большинстве

1.13

 karde

1.12

攥

1.10

 decorativo

1.10

 again

1.10

Activations Density 0.420%