INDEX

Explanations

words or phrases starting descriptions

The neuron activates on standalone numbers (numeric tokens) embedded in the text.

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

ĥ

-1.49

 drie

-1.48

доми

-1.44

戋

-1.43

return

-1.41

voted

-1.38

 Komun

-1.38

 Obten

-1.37

 cliche

-1.36

 citar

-1.35

POSITIVE LOGITS

of

2.31

</h2>

1.96

 that

1.95

 from

1.95

bánh

1.65

букет

1.60

asının

1.59

 likewise

1.55

which

1.51

</em>

1.50

Activations Density 0.000%