INDEX

Explanations

specific nouns

The neuron flags salient content nouns that denote methods, tools, solutions, products, services, or other key topic entities.

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

дают

-1.01

További

-1.00

-0.98

volent

-0.98

 esercit

-0.93

-0.91

Gemeinde

-0.90

ــ

-0.90

 sécher

-0.90

男孩

-0.88

POSITIVE LOGITS

 with

1.80

 that

1.68

ที่มี

1.45

 whose

1.30

 which

1.15

 Arquite

1.11

しそう

1.08

 Entrega

1.08

 Presidencia

1.06

Напомним

1.05

Activations Density 0.090%