INDEX

Explanations

recent and current time

The neuron flags temporal discourse markers—particularly adverbs and phrases that situate events in time (e.g. “now,” “nowadays,” “сейчас,” “almost immediately”).

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

 With

-1.84

 Before

-1.72

for

-1.70

了一個

-1.69

StreetMap

-1.69

as

-1.66

-1.62

of

-1.56

了一些

-1.56

 with

-1.55

POSITIVE LOGITS

뀨

1.68

 everything

1.64

久しぶりに

1.62

 affaires

1.61

最近の

1.56

ーズン

1.53

今年の

1.48

it

1.48

現在

1.45

今日も

1.43

Activations Density 0.013%