INDEX

Explanations

Uttar Pradesh

The neuron fires on recurrent, non–sentence-initial proper nouns—e.g. place-name components like “Uttar” or official titles like “Warden.”

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

-3.53

鶿

-2.52

but

-2.48

实的

-2.41

呓

-2.39

 drugim

-2.38

していました

-2.34

なんだけど

-2.30

 моих

-2.27

缡

-2.20

POSITIVE LOGITS

}$

3.02

2.98

大部分

2.80

 буквально

2.59

 However

2.55

史上

2.48

...?

2.47

自身

2.39

ции

2.38

가

2.25

Activations Density 0.012%