INDEX

Explanations

Feldmann names

The neuron fires on tokens containing the substring “feld,” e.g. names or place-names ending in “feld.”

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

-3.02

絰

-2.88

鱘

-2.72

);

-2.45

 放題

-2.36

-2.27

-2.25

许久

-2.16

 parachoque

-2.13

’;

-2.09

POSITIVE LOGITS

 mutta

2.81

家を

2.59

 большим

2.56

 settembre

2.53

If

2.52

。“

2.48

But

2.45

ANY

2.42

簡単な

2.28

 bapak

2.28

Activations Density 0.008%