INDEX

Explanations

end of Spanish words

The neuron responds strongly to numeric tokens and digit‐containing sequences (e.g. years, measurements, IDs, URLs).

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

 उसकी

1.25

નની

1.18

उनकी

1.08

 அவன்

1.02

 लेती

1.00

 அவனை

0.96

 उनकी

0.96

三个

0.95

 करती

0.93

 रखती

0.93

POSITIVE LOGITS

 PICTOGRAM

1.52

$\

1.36

-$\

1.34

 της

1.33

PICTOGRAM

1.29

 della

1.23

$\

1.18

($\

1.18

$_

1.17

$-\

1.17

Activations Density 0.083%