INDEX

Explanations

i.e. clarifies or defines

The neuron detects Latin-style parenthetical abbreviations like “i.e.” or “e.g.” used to introduce clarifications or examples.

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

fordshire

-1.76

-1.66

From

-1.55

ανα

-1.53

By

-1.52

My

-1.46

５

-1.46

до

-1.46

And

-1.46

ၜ

-1.45

POSITIVE LOGITS

 same

2.25

are

1.88

﹍﹍

1.72

🗹

1.69

 condena

1.63

𐬹

1.63

 mantend

1.57

 leichte

1.53

 vrea

1.49

 leichten

1.48

Activations Density 0.019%