INDEX

Explanations

year, date, country, place

The neuron detects tokens that are part of multi‐word named entities (proper names, technical terms, or specific titles).

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

以來

-1.89

باره

-1.43

 transférez

-1.42

を開催

-1.41

してました

-1.40

hommes

-1.38

してみた

-1.38

並不是

-1.38

 wesentlichen

-1.38

 Handlungen

-1.37

POSITIVE LOGITS

1.80

。

1.76

for

1.55

1.42

ifikation

1.42

jaar

1.41

тона

1.40

Producto

1.37

ária

1.34

 llam

1.34

Activations Density 0.308%