INDEX

Explanations

time leading up to events

The neuron detects temporal markers and scheduling cues—tokens that signal time boundaries or sequence (e.g. before, start, end, due, scheduled).

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

 навіть

-1.50

 ganska

-1.41

 varför

-1.36

vlád

-1.30

 nästan

-1.28

 utilisent

-1.27

Рецепт

-1.26

 різних

-1.26

Mengapa

-1.25

፲

-1.25

POSITIVE LOGITS

our

1.45

we

1.42

the

1.42

for

1.20

Our

1.20

 like

1.16

 considérons

1.16

для

1.08

 да

1.07

servlet

1.07

Activations Density 0.038%