INDEX

Explanations

case numbers with suffixes

The neuron is activated by numeric literals (tokens consisting of numbers and numeric constants).

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

呖

-1.70

 Hinblick

-1.69



-1.68

تیجه

-1.63

遢

-1.61

 dificil

-1.59

آیا

-1.56

!“

-1.55

 استفاده

-1.54

SpringRunner

-1.54

POSITIVE LOGITS

of

2.36

 also

1.90

1.89

in

1.84

</h2>

1.65

 pernikahan

1.55

 remaja

1.54

 again

1.51

</sup>

1.47

 撮り

1.45

Activations Density 0.001%