INDEX

Explanations

dimensions separating numbers

The neuron activates on the lowercase “x” (and adjacent numbers) when it’s used as the “by” separator in size or dimension specifications.

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

 consectetur

-1.23

its

-1.14

пиона

-1.11

lho

-1.06

 chemise

-1.02

🤯

-1.02

\""

-1.02

cended

-1.02

:')

-1.00

):-

-1.00

POSITIVE LOGITS

1.17

帰る

1.02

Suara

1.01

 jede

0.99

 pięk

0.97

Races

0.97

Facile

0.92

 இந்த

0.91

ürger

0.90

and

0.88

Activations Density 0.003%