Lesson 09 - Reading the Conjecture 3 Research Map

Prerequisites And Diagnostic Checks

Ask the learner to reconstruct the chain:

good QLDPC -> expansion -> cut demand -> hardware separator -> syndrome-depth lower bound.

Then ask which arrow is least clear. That answer should determine how much of this lesson becomes review versus research-map practice.

Concrete Motivation

The purpose of this course is not to memorize vocabulary. The purpose is to make the Conjecture 3 graph readable enough that you can contribute to it.

The research graph is large because it separates:

• theorem-backed claims
• conditional bridges
• open gaps
• boundary examples
• failed routes
• future frontier directions

This lesson teaches how to navigate that structure.

Tracing The Lower-Bound Proof Chain

Start with the current overview:

• ../../../../index.md
• ../../../../research/graph-audit.md

For the foundations course, the clean chain is:

1. Quantum Tanner codes provide theorem-level good QLDPC families.
2. Their construction is anchored in expander objects.
3. Expansion can force cross-cut syndrome demand.
4. Static 2D hardware has small balanced cuts.
5. Cross-cut demand divided by cut capacity gives syndrome-depth lower bounds.

This chain explains the already-established static 2D target.

Locating The Open Gaps

The frontier is not simply "prove a 2D lower bound." The static theorem-level target is already largely settled in the local graph.

The remaining gaps are sharper:

• transferring auxiliary quantum Tanner expansion to the exact stabilizer-presentation property needed by a route
• proving intrinsic balanced cut-rank lower bounds for explicit families
• giving the broad CD(T_n,G) object a compiler-native interpretation
• deciding whether ordinary graph routing, hypergraph routing, hidden-vertex graph cuts, or directly submodular demand is the right formal language

Congestion-Dilation Versus Stabilizer Cut Rank

Ordinary congestion-dilation thinks in terms of routing paths for edges of a guest graph. Stabilizer syndrome extraction can create demands that are better expressed as a cut-rank or submodular function.

The node stabilizer-cut-rank-functional.md defines an intrinsic quotient-rank quantity across a cut.

The node stabilizer-cut-rank-defines-canonical-submodular-cd-object.md packages that quantity as a canonical symmetric submodular demand object.

The node submodular-cut-congestion-lower-bounds-swap-only-compiler-depth.md shows that, once this submodular object is allowed, a SWAP-only cut-congestion lower bound is already theorem-level.

So the subtle remaining question is not whether some intrinsic demand object exists. It is whether it has the intended classical routing or compiler semantics.

Reading A Research Card Independently

Use this checklist for any node:

1. What is the claim?
2. Is it an established theorem, an inferred bridge, a conditional route, or a boundary marker?
3. What assumptions does it need?
4. Which earlier nodes does it depend on?
5. Which conjecture gap does it close?
6. Which gap remains open?
7. Does it change the teaching syllabus, or only the research frontier?

Do not read every node linearly. Start from the graph audit's canonical spine, then branch only when a node dependency matters.

Positioning Toward The Next Course

The next course should focus on the intrinsic side:

• stabilizer cut rank
• binary matroid connectivity
• branchwidth and tangles
• dense tangle breadth
• hidden-vertex graph-cut representability
• why ordinary routing language is not obviously enough

Those are the tools that enter after the foundations course.

What This Does And Does Not Prove

This lesson does not close Conjecture 3. It turns the research map into a working navigation system.

The most important discipline is to keep theorem, intuition, bridge, and open gap separate. Much of the confusion in this project disappears when those labels are kept explicit.

Active Recall

• Which part of the static 2D target is already theorem-backed?
• What is the anchor gap for quantum Tanner codes?
• Why is stabilizer cut rank more intrinsic than a chosen Tanner presentation?
• Why does a submodular demand object not automatically give ordinary packet-routing semantics?
• What should the next advanced course teach?

Next-Step Handoff

After this lesson, the recommended next course is a matroid-and-cut-rank frontier course. It should start from ../../../../research/graph-audit.md, not from a generic matroid textbook.