INDEX

Explanations

Michael Cohen, Q:

The neuron strongly activates on the standalone “.” punctuation token, particularly when used as the module‐separator in Python import statements (e.g. “from .api import …”).

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

Ꭳ

-1.96

 Informações

-1.69

 cenu

-1.60

鏐

-1.59

 Comprometido

-1.55

 erfolgreiche

-1.50

 Após

-1.48

蛩

-1.47

 płat

-1.47

 Pentru

-1.47

POSITIVE LOGITS

</strong>

2.45

เหมือน

1.88

虽然

1.85

–

1.79

あ

1.67

<td>

1.66

</i>

1.63

雖然

1.63

</h3>

1.59

年来

1.59

Activations Density 0.016%