MRH Explorer
Explore the Markov Relevancy Horizon — the context boundary that defines what each agent can see and affect. Adjust horizon depth, explore the 4D MRH profile, and see how distance decays trust.
Horizon Depth2 hops
Only selfDirect (1)Default (2)Max (3)Beyond
Explore Nodes (hover to highlight)
5 entities beyond horizon (invisible)
4D Horizon Profile
Each action has a 4-dimensional context that determines its cost, required trust, and visibility.
Spatial (ΔR)40% weight
Temporal (ΔT)30% weight
Complexity (ΔC)30% weight
Quality (ΔQ)0% weight
ATP Cost
15 base x 1.5x quality
23 ATP
Quick Presets
Why Context Boundaries?
In the real world, you can't see everything. You know your friends, have some sense of their friends, and beyond that it's mostly unknown. MRH formalizes this natural limitation as a privacy-preserving context boundary.
Natural Privacy
Entities beyond your horizon can't see your relationships. Privacy emerges from network structure, not access control lists.
Computational Efficiency
With ~10 connections per entity, depth-3 traversal covers ~1,000 entities instead of the entire network. O(d³) vs O(n).
Spam Prevention
Actions outside your horizon cost exponentially more ATP. You can't spam people you don't know without burning through resources.
Trust Accuracy
Beyond 3 hops, trust signals degrade below statistical significance (Markov property). MRH stops before trust becomes noise.
The key insight: MRH creates a world where privacy, efficiency, and accuracy all align. You see what's relevant, can't spy on what isn't, and the system naturally scales without central coordination.
Interactive Tools
View all tools →Also explore