INDEX

Explanations

WithIdentifier code identifiers

The neuron fires most strongly on demonstrative pronouns meaning “this/these” (e.g. German diese, Czech tato) across different languages.

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

ѣмъ

-2.86

樒

-2.47

"—

-2.42

celotti

-2.33

埽

-2.30

を始め

-2.28

不该

-2.27

-2.19

');

-2.14

縢

-2.11

POSITIVE LOGITS

2.39

</strong>

2.28

</h3>

2.22

谖

2.22

 ที่

2.20

vantaged

2.16

})

2.11

 kinds

2.11

一个

2.06

 sélection

2.06

Activations Density 0.000%