INDEX

Explanations

text-align center

The neuron is triggered by the word “initial” (as in the “initial-scale” setting of a viewport meta tag).

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

_{

-1.66

↵↵↵

-1.58

</h4>

-1.55

↵↵↵↵

-1.46

↵↵↵↵↵

-1.44

($__

-1.44

</h5>

-1.43

↵↵↵↵↵↵

-1.38

}{

-1.33

-1.32

POSITIVE LOGITS

 Männern

1.98

 Однако

1.73

 dringend

1.72

 Новости

1.58

 Seventy

1.56

 damski

1.55

 Mitgliedern

1.55

ahme

1.52

 Namun

1.48

 Theologie

1.48

Activations Density 0.009%