Chapter 16 — Open Problems

Prerequisites: Parts I--V.

This chapter consolidates all open problems and conjectures from throughout the theory.

16.1 Conjecture 7.10: Gauge-Fixing Uniqueness (Non-Abelian $G$)#

Statement#

If $h_1^*$ and $h_2^*$ are two global minima of $\mathcal{E}_{F^\circ}$ with $h_1^*\not\sim_{\mathcal{G}_{\mathrm{const}}}h_2^*$, then:

Status#

Approaches#

1. Tree graphs: no loops, so fixing a root determines all other gauges uniquely (1-variable convex optimisation at each step).

2. $SU(2)$ Morse theory: classify critical points of $\mathcal{E}_{F^\circ}:G^{F^\circ}\to\mathbb{R}$ by Morse index. Show that all global minima lie on a single $\mathcal{G}_{\mathrm{const}}$-orbit.

3. Bypass strategy: even if the conjecture fails, the sequential protocol (Construction 7.6) ensures $\Sigma$ and $\mathbf{e}$ are well-defined. The issue affects only the reading of $[A_\infty]$ via $\rho$. Using holonomy invariants (conjugacy classes) instead of $\rho$ sidesteps the conjecture entirely.

16.2 Fruit Dynamics and Continuity#

Problem#

Define rigorously when $F_t\in\mathfrak{F}_t$ and $F_{t'}\in\mathfrak{F}_{t'}$ represent "the same existence evolving".

Proposed definitions (Chapter 14)#

• Symmetric-difference criterion: $|F_t\triangle F_{t'}|/|F_t|\le\epsilon$.
• Moduli distance: $d_{\mathcal{M}}(\mathrm{Ex}_t(F_t),\mathrm{Ex}_{t'}(F_{t'}))\le\delta$.

Open sub-problems#

1. Lifespan: define and compute the duration over which a fruit persists.
2. Splitting/merging: classify the events where one fruit becomes two (or vice versa), including the topological signature of the transition.
3. Door pattern evolution: how does $\Sigma_t$ change as the relational field evolves?

16.3 Gauge--Holonomy Coupling for Doors#

Conjecture#

Energy-based doors and holonomy-based doors are related:

That is, a holonomy anomaly always accompanies an energy anomaly.

Evidence#

• For $G=U(1)$: holonomy anomaly at node $i$ requires non-zero external coupling through $i$, so the conjecture holds with $\tau'=\tau/C$ for some constant $C$.
• Converse fails in general: large $b_{F,t}(i)$ does not imply large holonomy anomaly (a flat connection with large leakage is possible).

16.4 Inter-Fruit Interaction (Meta-Relational Field)#

Problem#

Describe the interaction between two fruits $F_1,F_2$ in a principled way.

Proposal#

Define a meta-relational field with:

• Meta-weight: $W^{\mathrm{meta}}(F_1,F_2):=\sum_{i\in F_1}\sum_{j\in F_2}W_t(i,j)$.
• Meta-transit: an appropriately averaged holonomy between $F_1$ and $F_2$.

Open sub-problems#

1. Well-definedness of the averaged holonomy (path dependence for non-abelian $G$).
2. Recursive structure: do "meta-fruits" exist in the meta-relational field?
3. Information loss quantification: how much structure is lost in coarse-graining?

16.5 Extension to Infinite $V$#

Problem#

What survives as $|V|\to\infty$?

Directions#

1. Thermodynamic limit: scaling of fruit size $|F_n|/n$, conductance $\phi(F_n)$, and door density $|\Sigma_n|/|F_n|$.

2. Continuum limit: under appropriate scaling, the discrete relational field should converge to a continuous gauge connection, with:

   • Discrete conductance $\to$ Cheeger constant.
   • Discrete holonomy $\to$ continuous curvature.
   • Discrete doors $\to$ Uhlenbeck singular points.

3. Graphon limit: the sequence of weighted graphs converges to a graphon $W:[0,1]^2\to\mathbb{R}_{\ge0}$. Extending the gauge structure to graphons requires a non-trivial construction.

16.6 Information-Theoretic Interpretation#

Conjecture#

The mutual information between a fruit and its exterior is bounded by door energy:

Motivation#

Doors are the only channels of information between a fruit and the exterior (by the no-boundary principle). The door energy $e_p$ quantifies the bandwidth of channel $p$.

16.7 Computational Complexity#

Known#

• Finding the minimum-conductance subset is NP-hard in general.
• Spectral methods give $O(\sqrt{\theta})$-approximation (discrete Cheeger inequality).
• For $G=U(1)$, optimal gauge is a linear system (polynomial time).

Open#

• For general compact $G$: is finding the global minimum of $\mathcal{E}_{F^\circ}$ NP-hard?
• Can the number of local minima of $\mathcal{E}_{F^\circ}$ for $G=SU(2)$ be bounded in terms of graph invariants?
• Efficient algorithms for Axes 1--3 computation on large graphs ($|F|>10^4$).

16.8 Non-Abelian Convergence Rate#

Problem (from Chapter 13)#

For $G=SU(2)$, determine the Lojasiewicz exponent $\alpha$ at critical points of $\mathcal{E}$. This determines whether convergence of the YM flow is exponential ($\alpha=1/2$) or polynomial ($\alpha>1/2$).

Related#

• Classify the Morse--Bott structure of $\mathcal{E}$ on $G^{|E|}$ for specific graph families.
• Determine whether Gribov copies are isolated or form continuous families.

16.9 Yang--Mills Uniqueness for Non-Abelian Groups (SU(2))#

Problem#

For a fixed graph topology $(n,E)$ and fruit interior $F^\circ$, the flattening energy $\mathcal{E}_{F^\circ}(h) = \sum_{\triangle} \|\Omega(\triangle)\|_G^2$ may have multiple global minima when $G = \mathrm{SU}(2)$.

Question: Is the minimizer of $\mathcal{E}_{F^\circ}(h)$ unique (up to gauge equivalence of type $\mathcal{G}_{\mathrm{const}}$)? If not, how does non-uniqueness affect the definition of the existence triple $([A_\infty], \Sigma, \mathbf{e})$ and the state space $C(n,E)$?

Implications#

• If YES (unique): The canonical connection $[A_\infty]$ is well-defined, and the state space $C(n,E) = P(n,E) / G^n$ is a quotient manifold (possibly with orbifold singularities only at flat configurations).
• If NO (non-unique): The state space has genuine reducibility at certain configurations. The existence triple becomes multi-valued: $\mathrm{Existence}(F,t) = ([A_\infty]^{(k)}, \Sigma, \mathbf{e})$ for $k=1,\ldots,m$ distinct minimizers. This complicates the semantic interpretation of the world.

Related Literature#

• Connes--Douglas (1995): Yang--Mills minima for SU(2) connections on compact manifolds.
• Donaldson (1990): Gauge theory on 4-manifolds; instanton moduli spaces.
• Discrete setting: Rade (1994) on discrete YM energy, but non-uniqueness not fully addressed.

16.10 Event Completeness and Minimality#

Problem#

The six-event vocabulary $\{\mathrm{DEFORM}, \mathrm{CONTACT}, \mathrm{BIRTH}, \mathrm{DEATH}, \mathrm{SPLIT}, \mathrm{MERGE}\}$ is motivated by physical intuition.

Questions:

1. Are these six events complete? Can every admissible world trajectory be uniquely decomposed into a sequence of these events?
2. Are they minimal? Is there a smaller set from which all others are derived?
3. Are they mutually exclusive? Can two events fire simultaneously, or must they be serialized?

Evidence and Sub-Problems#

1. Primitive vs. derived: SPLIT and MERGE are detected threshold-crossings (derived from continuous DEFORM), while BIRTH/DEATH/CONTACT are primitive (true topology changes). This suggests a refinement:

   • Primitive events: DEFORM, CONTACT, BIRTH, DEATH (4 events).
   • Detected events: SPLIT, MERGE (consequences of DEFORM crossing a threshold).
   • Synchronous firing: Can BIRTH and CONTACT fire at the same time? (A new node might immediately CONTACT existing nodes.)

2. Commutativity: Do the event sequences $(e_1, e_2)$ and $(e_2, e_1)$ always lead to the same final state (up to gauge equivalence)? This is crucial for the evidence log to have a canonical interpretation.

3. Reversibility: For each event $e$, does an inverse event $e^{-1}$ exist? (E.g., is CONTACT reversible by a DEFORM that drives the edge weight to zero?)

16.11 Inter-Stratum Topology and Continuity#

Problem#

The total state space $\mathcal{S} = \bigsqcup_{(n,E) \in \mathrm{Top}} C(n,E)$ is a disjoint union of spaces with different dimensions.

Question: Is there a natural (e.g., metric, diffeological, or compactification) topology on $\mathcal{S}$ such that:

1. Events like BIRTH and DEATH induce continuous maps (or at least measurable in a meaningful sense)?
2. The stratification respects the topology (strata are open/closed sets)?
3. Sequences of configurations from different strata can converge in a well-defined sense?

Motivation#

In standard dynamical systems, the state space is a fixed manifold, and evolution is continuous. In RelationWorld, topology changes are discontinuous. Understanding the "nearness" of states in different topologies would allow us to define:

• Limit points of configuration sequences.
• Compactifications of $\mathcal{S}$ (adding "boundary" states corresponding to infinite growth or collapse).
• Continuity of the global transition map.

Technical Directions#

1. Quotient topology: Equip $\mathcal{S}$ with the quotient topology from a larger space (e.g., all finite node sets, all edge sets).
2. Diophantine geometry: Use algebraic methods to define "distance" between topologies (e.g., Gromov--Hausdorff).
3. Diffeological spaces (Frölicher--Kriegl): allow a more flexible notion of smoothness across strata.

16.12 Evidence Completeness and Invertibility#

Problem#

The evidence extraction functional:

captures what changed. But can it certify why a particular event occurred?

Questions:

1. Completeness: Is $\mathrm{Ev}(\cdot)$ sufficient to uniquely reconstruct the event $e_t$ from the state transition?
2. Invertibility: Given the evidence log $\{\mathrm{Ev}(s_i, s_{i+1}, e_i)\}_{i=0}^{T}$, can we recover the full trajectory $\{s_i\}$ (up to gauge equivalence)?
3. Disambiguation: If two events would produce the same $\mathrm{Ev}$ output, how do we distinguish them?

Example#

Suppose a single edge $(i,j)$ weight decreases dramatically. Two interpretations:

• DEFORM explanation: a smooth decrease via the DEFORM mechanism.
• SPLIT consequence explanation: the decrease triggered a SPLIT, which is then the "true" event.

The evidence log $(e_t, \Delta W, \Delta \Omega, \Delta \mathfrak{F}, \Delta \Sigma)$ includes $\Delta \mathfrak{F}$ (fruit changes), which disambiguates. But formally proving completeness requires axiomatizing the relationship between events and evidence.

16.13 Fruit Structure as a Continuous Functor#

Problem#

The fruit set $\mathfrak{F}_t$ is derived from the weights $W_t$ via the Cheeger conductance:

Questions:

1. Does $\mathfrak{F}_t$ vary continuously with $W_t$ in a well-defined topological sense?
2. Can the fruit-detection algorithm be viewed as a functor $\Phi$ from the category of weighted graphs to the category of simplicial complexes?
3. What are the continuity properties of $\Phi$? For instance, does small perturbation of $W_t$ lead to small change in $\mathfrak{F}_t$?

Technical Context#

Theorem F (Chapter 8) guarantees spectral stability: small changes in $W_t$ lead to small changes in individual conductances $\phi_t(F)$. However, this does not immediately imply continuity of the fruit set:

• A bifurcation at $\phi_t(F) = \theta$ is discontinuous (a fruit appears/disappears).
• Two fruits may merge at $\phi_t(F_1 \cup F_2) = \theta$, creating a discontinuity.

Applications#

• Stability of Axes: If $\mathfrak{F}_t$ is continuous, then the cohomological Axes 1--3 (Chapter 12) vary continuously, and their persistence barcodes have well-defined limits.
• Persistence theory: The fruit set can be organized into a persistence diagram showing the "lifespan" of each fruit as parameters vary.

16.14 Minimality and Sufficiency of State Variables#

Problem#

The raw state $(V, w_t, g_t)$ (or quotient $(n,E, [p])$ where $[p] \in C(n,E)$) contains potentially redundant information.

Questions:

1. Is $(w_{ij}, g_{ij})$ per edge the minimal sufficient statistic for the world, or can the system be described more parsimoniously?
2. Specifically, do the existence triples $\{\mathrm{Existence}(F,t)\}_{F \in \mathfrak{F}_t}$ fully determine the relational field, up to gauge equivalence?
3. Can nodes be eliminated from the description, leaving only the fruits (higher-order structures) as primitives?

Evidence#

For sufficiency of $(w, g)$: The world $\mathfrak{W}_t = (\mathcal{W}_t, \mathfrak{F}_t, \Sigma_t)$ is fully determined by the relational field $\mathcal{W}_t$, which in turn is determined by the gauge class $[p]$ of $(w, g)$ (Theorem 8.4).

Against minimality of $(w, g)$: The existence triples contain the "essential" information about each fruit's internal structure and boundary conditions. In a coarse-grained (multi-scale) description, specifying only the existence triples per fruit might suffice, with inter-fruit couplings captured by the inter-fruit weights and transits.

Research Directions#

1. Coarse-graining: Develop a formal procedure to "integrate out" nodes within each fruit, leaving a reduced system of $|\mathfrak{F}_t|$ meta-nodes.
2. Emergence: Investigate whether the world structure (fruits, doors, Axes) emerges from a lower-level description (e.g., algebraic structures on the universal cover).
3. Quantum analogue: If a quantum analogue of RelationWorld is formulated (see Chapter 15), does the state space reduce to eigenspaces of certain operators?