INDEX

Explanations

borders and directions

The neuron detects mentions of geographic boundary language—tokens like “borders,” “bordered,” direction words and adjoining place names that describe what a region is adjacent to.

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

 prostředí

-1.21

 сайті

-1.10

 piña

-1.09

原始内容存档于

-1.07

 ͡°)

-1.06

 違い

-1.05

uebas

-1.04

 سينما

-1.04

 biografias

-1.03

 strategy

-1.02

POSITIVE LOGITS

mar

1.13

fea

1.11

few

1.07

Å

1.05

 edged

1.05

1.02

0.99

rrrrrr

0.98

conne

0.97

ra

0.96

Activations Density 0.025%