INDEX

Explanations

débat, début, decor

The neuron activates on tokens containing accented characters (especially French-style “é”).

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

-3.66

lgari

-2.89

which

-2.78

this

-2.59

渽

-2.52

its

-2.48

 Reaktion

-2.39

ーズン

-2.36

доры

-2.36

↵↵

-2.27

POSITIVE LOGITS

’

3.39

鲥

2.77

了一种

2.52

zelfde

2.50

),

2.44

了一個

2.39

涐

2.25

anleitung

2.22

 apparaat

2.22

鲔

2.13

Activations Density 0.001%