INDEX

Explanations

word puzzles

The neuron activates on proper nouns referring to specific products, brands, or item names.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ERCISE

0.77

INITIAL

0.74

ISION

0.72

TABLE

0.71

 それぞれ

0.70

 PLASMA

0.70

GOING

0.68

VALID

0.68

 Ongoing

0.68

UAGES

0.67

POSITIVE LOGITS

 çünkü

0.65

 murió

0.64

is

0.64

isn

0.64

сит

0.64

cra

0.63

vì

0.63

怎么

0.63

 fordi

0.63

vt

0.62

Activations Density 0.019%