INDEX

Explanations

days, months, or starters

The neuron specifically detects occurrences of the uppercase keyword “ACTUAL.”

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

 prenota

-2.38

adeira

-2.30

他

-2.28

㈬

-2.27

 土産

-2.25

They

-2.23

媖

-2.23

葙

-2.22

 météoro

-2.22

 pédagog

-2.22

POSITIVE LOGITS

</h2>

3.05

2.69

 أن

2.63

2.59

—

2.58

</em>

2.50

<strong>

2.47

氿

2.33

</h5>

2.33

Are

2.30

Activations Density 0.000%