INDEX

Explanations

drawing rectangles

This neuron identifies mentions of geometric shape names and related spatial‐geometry terms.

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

 gebeten

-1.45

США

-1.45

 inalámbrica

-1.38

umumkan

-1.34

 ไทย

-1.34

 sosok

-1.32

Ⓣ

-1.30

tapkan

-1.29

 starte

-1.29

ÇÕES

-1.28

POSITIVE LOGITS

our

1.68

千萬

1.32

"};

1.30

遄

1.27

 others

1.25

})}

1.23

 dirigez

1.23

</h5>

1.23

ومة

1.23

 more

1.21

Activations Density 0.038%