INDEX

Explanations

day and night

The neuron detects explicit mentions of times of day (e.g. “night,” “day,” “nighttime,” “daytime”).

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

 verschillende

-1.39

 kvalitet

-1.34

 migliorare

-1.27

を買う

-1.27

 dreapta

-1.27

 lembra

-1.26

Menge

-1.25

Rasp

-1.24

 livello

-1.24

sach

-1.23

POSITIVE LOGITS

 when

1.44

 With

1.41

The

1.41

 This

1.37

 When

1.35

1.30

In

1.29

</i>

1.28

1.27

 These

1.27

Activations Density 0.021%