INDEX

Explanations

java array declaration

The neuron fires on the array‐declaration brackets (the “[]” tokens) in code.

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

Of

-1.91

If

-1.82

There

-1.81

or

-1.72

To

-1.70

We

-1.69

 erscheinen

-1.69

whats

-1.62

cannot

-1.61

おわりに

-1.61

POSITIVE LOGITS

一世

1.70

五行

1.70

</h3>

1.68

 PUSTAKA

1.63

ſy

1.62

腹部

1.54

胚

1.50

뻗

1.48

琥

1.47

 ledig

1.45

Activations Density 0.007%