INDEX

Explanations

Check your policy

The neuron activates on bureaucratic or procedural language—especially words describing planning, preparation, outlines, stages, or processes.

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

SpriteCollision

0.75

 ফিল্ম

0.70

面白

0.68

ഠ

0.67

Ironically

0.66

面白い

0.64

coeffs

0.63

 Appreciation

0.63

 желательно

0.63

 perturbative

0.63

POSITIVE LOGITS

 FAQs

1.66

FAQ

1.41

FAQ

1.38

FAQs

1.34

faq

1.29

 Frequently

1.18

faq

1.08

Frequently

1.03

 frequently

1.00

 guide

0.97

Activations Density 0.610%