INDEX

Explanations

area around locations

The neuron fires on tokens that express spatial relationships or locational context (e.g. words like “lying,” “underneath,” “adjacent,” “feet,” “immediately”) indicating position or distance.

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

 also

-1.73

for

-1.28

istiques

-1.22

('');

-1.21

autoreleasepool

-1.20

 find

-1.19

 cilt

-1.17

’，

-1.16

 during

-1.16

存知

-1.14

POSITIVE LOGITS

you

1.75

 jednej

1.47

 могут

1.42

including

1.37

してみてください

1.36

 pensez

1.34

してくれた

1.34

 яку

1.34

 zweiter

1.32

 лучше

1.29

Activations Density 0.066%