INDEX

Explanations

dates and times

The neuron detects date‐ and time‐related tokens (e.g., month names, numeric dates, times, time zones, and colons).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

if

-2.22

}$,

-1.61

 When

-1.50

 only

-1.50

 setMessage

-1.49

四年

-1.48

,”

-1.48

adget

-1.47

鏖

-1.45

 people

-1.45

POSITIVE LOGITS

we

1.70

1.54

ay

1.50

and

1.45

だけでなく

1.43

of

1.36

ity

1.32

遠く

1.31

 mentales

1.27

Архівовано

1.27

Activations Density 0.022%