numbers following symbols

The neuron is effectively a “dead” neuron here—it never activates on any tokens in these samples.

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

0.20

yrs

0.19

0.18

0.17

 July

0.17

৬

0.17

 volts

0.17

POSITIVE LOGITS

авто

0.18

 discursive

0.17

Список

0.17

他和

0.17

を行う

0.16

 schemas

0.16

对方

0.16

പരമായ

0.16

व्यक्ति

0.16

प्रदर्शन

0.16

Activations Density 0.007%