INDEX

Explanations

that, which introducing relative clauses

The neuron is triggered by multi‐syllabic, domain‐specific content words—especially technical nouns used in academic or scientific descriptions.

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

 में

-1.34

ponenten

-1.27

is

-1.20

Și

-1.16

 دارای

-1.16

}')

-1.14

”，

-1.13

有

-1.13

كويكب

-1.13

 способы

-1.13

POSITIVE LOGITS

 that

1.61

our

1.52

 this

1.52

 every

1.26

 these

1.25

so

1.25

 those

1.25

 accompany

1.23

 From

1.20

 make

1.18

Activations Density 0.047%