INDEX

Explanations

names, brands, and product descriptions

The neuron is highly activated by typographic apostrophes (’ or ') appearing in words or possessives.

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

 olaca

-0.86

ورات

-0.83

racha

-0.81

pesar

-0.80

合衆

-0.79

撰文

-0.77

ỷ

-0.76

انو

-0.75

âteau

-0.73

 Phòng

-0.73

POSITIVE LOGITS

 Piaget

0.84

 eres

0.78

 lignes

0.76

Clou

0.75

 Nacho

0.74

Interpret

0.73

LINE

0.73

 gire

0.72

 complication

0.71

zial

0.70

Activations Density 0.014%