INDEX

Explanations

this year

This neuron fires on the introductory metadata and time‐marker tokens that begin new sections—e.g. dates, “This year,” “This month,” headings, and similar post or article lead-ins.

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

 these

-1.59

 this

-1.51

 when

-1.36

 posled

-1.33

صاد

-1.18

 ONLY

-1.18

 days

-1.16

 here

-1.16

 usarlo

-1.15

 atentos

-1.14

POSITIVE LOGITS

вать

1.18

 другие

1.16

 مدى

1.12

 जिसमें

1.12

</h1>

1.10

 więks

1.09

1.07

กับ

1.05

자가

1.05

‘

1.03

Activations Density 0.105%