INDEX

Explanations

graph edges with connections

This neuron detects mentions of graph “edges” or edge‐related relationships.

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

gradable

-1.52

 visage

-1.40

 Hilo

-1.38

初め

-1.36

 chanteuse

-1.35

 taxe

-1.35

»;

-1.35

agonal

-1.34

-1.33

“

-1.32

POSITIVE LOGITS

 olika

1.70

ݯ

1.67

the

1.59

፨

1.53

totta

1.45

pherds

1.41

 ก็

1.41

 ESTADO

1.41

 различ

1.40

 distintas

1.39

Activations Density 0.034%