INDEX

Explanations

characters and figures

This neuron responds to words and phrases that signal major plot developments or spoilers—terms like “plot twist,” “happens,” “culminates,” and “find out.”

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

 market

0.91

 बाजार

0.81

 reseller

0.81

 specialized

0.80

 agribusiness

0.79

 haven

0.78

 aziende

0.78

 servicing

0.77

专业的

0.77

 factories

0.76

POSITIVE LOGITS

 సన్నివేశ

1.43

 сюжет

1.42

剧情

1.40

 subplot

1.39

劇情

1.29

 storyline

1.28

 эпизо

1.26

 ฉาก

1.26

 장면

1.23

主人公

1.21

Activations Density 0.541%