INDEX

Explanations

Without

The neuron detects words or morphemes expressing absence—especially the word “without” and the “-less” suffix.

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

 Bereichen

-1.55

你喜欢

-1.48

龔

-1.43



-1.41

タブ

-1.40

琑

-1.38

 pensado

-1.38

 verkau

-1.36

ριο

-1.30

 працю

-1.29

POSITIVE LOGITS

1.83

1.58

見ても

1.47

1.45

^{\

1.43

 ihres

1.41

 ihrem

1.40

1.34

 would

1.34

(\

1.34

Activations Density 0.012%