Research-Stack/6-Documentation/docs/research/unsolved_hard_problems_leverage_points.md

RRC Unsolved-Problems Leverage Analysis

Source: 6-Documentation/docs/research/unsolved_hard_problems_rrc_alignments.json
Schema: unsolved_hard_problems_rrc_alignments_v1 (68 problems, 67 unsolved, 1 solved boundary)
Generated: 2026-06-20

1. Methodology

I modeled the artifact as a directed influence graph:

• Nodes: the 67 unsolved problems (the solved 3D Poincaré boundary marker was excluded).
• Edge weights were assembled from two sources:
  1. The crossing_matrix row for each problem (0–1 connection strengths).
  2. known_reductions_to / known_reductions_from lists: a known reduction between A and B adds a 0.25 machinery bonus to the corresponding directed edge, because solving the stronger statement typically collapses or heavily informs the weaker one.
• Direction: an edge X → Y means "solving X is expected to unlock or collapse Y" via reduction, shared machinery, or analogy.

Leverage scoring

For each problem I computed four normalized components and combined them into a composite leverage score:

Component	Weight	Meaning
Strong weighted out-degree (edges ≥ 0.5)	30 %	Direct, high-confidence unlock pressure
Strong transitive reach (edges ≥ 0.5)	25 %	Number of problems reachable through rigorous-looking reductions / machinery
Two-hop strong reach	20 %	Immediate cascade depth before saturation
Cross-cluster bridging (strong edges)	15 %	Number of distinct alignment clusters unlocked outside the node's home cluster
Strong degree count	10 %	Count of direct high-confidence outgoing edges

I also ran a second pass with a 0.3 threshold to capture shared-technique / analogical links. The results below distinguish the two whenever the difference matters.

2. Ranked Top-10 Leverage Problems

The scoring is dominated by the computational-complexity core: P vs NP and its satellites act as the artifact's central routing hub. Only one non-complexity problem (Birch–Swinnerton-Dyer) cracks the extended top 15.

Rank	Problem	ID	Composite score	Strong reach	Cross clusters (strong)	RRC shape	RRC status	Home cluster
1	P vs NP	p_vs_np	0.975	11	2	CognitiveLoadField	CANDIDATE	cluster_02 (Complexity)
2	Unique Games Conjecture	unique_games_conjecture	0.900	11	2	CognitiveLoadField	CANDIDATE	cluster_02
3	Exponential Time Hypothesis	exponential_time_hypothesis	0.890	11	2	CognitiveLoadField	CANDIDATE	cluster_02
4	BPP vs P (derandomization)	bpp_vs_p	0.885	11	2	CognitiveLoadField	HOLD	cluster_02
5	Strong Exponential Time Hypothesis	strong_exponential_time_hypothesis	0.871	11	2	CognitiveLoadField	CANDIDATE	cluster_02
6	Discrete logarithm in P	discrete_log_in_p	0.867	11	2	CognitiveLoadField	CANDIDATE	cluster_02
7	Existence of NP-intermediate problems	np_intermediate_existence	0.846	11	2	CognitiveLoadField	HOLD	cluster_02
8	Integer factorization in P	factoring_in_p	0.829	11	2	CognitiveLoadField	CANDIDATE	cluster_02
9	Graph isomorphism in P?	graph_isomorphism_in_p	0.808	11	2	CognitiveLoadField	CANDIDATE	cluster_02
10	Quantum supremacy verification	quantum_supremacy_verification	0.617	11	2	ComputeKernelReceipt	HOLD	cluster_07 (Quantum/information)

Details per top problem

1. P vs NP (p_vs_np)

   • Home cluster: cluster_02 (Computational complexity core); the artifact also lists it in cluster_06 (Logic/foundations).
   • Clusters connected (strong): cluster_02, cluster_03 (PDE/field singularities), cluster_07 (quantum/information).
   • Direct strong unlocks: BPP vs P, discrete log, ETH, factoring, graph isomorphism, NP-intermediate existence, quantum supremacy verification, SETH, UGC.
   • With analogies (≥0.3): also reaches continuum hypothesis, consistency of ZFC, cosmological constant, dark matter, Navier–Stokes, Hilbert's 6th/16th, Yang–Mills mass gap (via quantum supremacy).
   • Why it tops the list: highest strong weighted out-degree (7.20) and the broadest set of direct strong edges in the artifact.

2. Unique Games Conjecture (unique_games_conjecture)

   • Home cluster: cluster_02.
   • Clusters connected (strong): cluster_02, cluster_03, cluster_07.
   • Direct strong unlocks: BPP vs P, discrete log, ETH, factoring, graph isomorphism, NP-intermediate, P vs NP, small-set expansion, SETH.
   • Note: UGC and small-set expansion are mutually reducible in the artifact; UGC is the more central hub.

3. Exponential Time Hypothesis (exponential_time_hypothesis)

   • Home cluster: cluster_02.
   • Clusters connected (strong): cluster_02, cluster_03, cluster_07.
   • Direct strong unlocks: BPP vs P, discrete log, factoring, graph isomorphism, NP-intermediate, P vs NP, SETH, UGC.
   • Role: fine-grained-complexity anchor; its collapse propagates through the entire complexity web.

4. BPP vs P (derandomization) (bpp_vs_p)

   • Home cluster: cluster_02.
   • Clusters connected (strong): cluster_02, cluster_03, cluster_07.
   • Direct strong unlocks: discrete log, ETH, factoring, graph isomorphism, NP-intermediate existence, P vs NP, quantum supremacy, SETH, UGC.
   • Status: HOLD (the artifact flags derandomization as currently blocked by circuit-lower-bound barriers).

5. Strong Exponential Time Hypothesis (strong_exponential_time_hypothesis)

   • Home cluster: cluster_02.
   • Clusters connected (strong): cluster_02, cluster_03, cluster_07.
   • Direct strong unlocks: BPP vs P, discrete log, ETH, factoring, graph isomorphism, NP-intermediate, P vs NP, UGC.
   • Relationship: tightly coupled to ETH (1.30 mutual edge), so solving either collapses the other and the rest of the cluster.

6. Discrete logarithm in P (discrete_log_in_p)

   • Home cluster: cluster_02.
   • Clusters connected (strong): cluster_02, cluster_03, cluster_07.
   • Direct strong unlocks: BPP vs P, ETH, factoring, graph isomorphism, NP-intermediate, P vs NP, SETH, UGC.
   • Note: mutually reducible with factoring; together they are the cryptographic hardness sub-hub.

7. Existence of NP-intermediate problems (np_intermediate_existence)

   • Home cluster: cluster_02.
   • Clusters connected (strong): cluster_02, cluster_03, cluster_07.
   • Direct strong unlocks: BPP vs P, discrete log, ETH, factoring, graph isomorphism, P vs NP, SETH, UGC.
   • Status: HOLD — the artifact notes it is conditional on P ≠ NP, so it is essentially a corollary-shaped gate.

8. Integer factorization in P (factoring_in_p)

   • Home cluster: cluster_02.
   • Clusters connected (strong): cluster_02, cluster_03, cluster_07.
   • Direct strong unlocks: BPP vs P, discrete log, ETH, NP-intermediate, P vs NP, SETH, UGC.
   • Note: like discrete log, a concrete algorithmic collapse node rather than a purely logical implication.

9. Graph isomorphism in P? (graph_isomorphism_in_p)

   • Home cluster: cluster_02.
   • Clusters connected (strong): cluster_02, cluster_03, cluster_07.
   • Direct strong unlocks: BPP vs P, discrete log, ETH, NP-intermediate, P vs NP, SETH, UGC.
   • Distinctive feature: its quasipolynomial witness gives it the highest proof_readiness (0.30) among the top complexity nodes.

10. Quantum supremacy verification (quantum_supremacy_verification)

    • Home cluster: cluster_07 (Quantum and information).
    • Clusters connected (strong): cluster_02, cluster_03.
    • Direct strong unlocks: P vs NP, Yang–Mills mass gap.
    • Why it ranks here despite fewer direct edges: it sits in cluster_07 and has high cross-cluster bridging; from it the strong transitive closure reaches the full complexity core and PDE/field-singularity nodes.
    • Status: HOLD — the artifact treats the verification gap as underspecified.

3. Cluster-Specific Leverage Leaders

Because the global top 10 is almost entirely the complexity core, the non-complexity clusters each have their own local leverage nodes. These are the best "entry points" if the goal is to collapse a particular domain rather than the whole graph.

Cluster	Best leverage node	Score	Strong reach	Key unlocks
cluster_01 — Millennium, L-functions, motives	Riemann Hypothesis	0.289	4	GRH, Birch–Swinnerton-Dyer, Hodge, twin primes
cluster_03 — PDE regularity/singularities	Hilbert's 16th problem	0.165	2	Navier–Stokes existence, Yang–Mills mass gap
cluster_04 — Arithmetic/Diophantine	Twin prime conjecture	0.312	4	Polignac, Elliott–Halberstam, Schinzel H, Goldbach
cluster_05 — Topology/geometry	Smooth 4D Poincaré conjecture	0.090	1	Generalized Poincaré (smooth)
cluster_08 — Cosmology/dark sectors	Dark energy equation of state	0.199	3	Cosmological constant, dark matter, inflation
cluster_09 — Fluid/field singularities	Navier–Stokes existence and smoothness	0.197	2	Navier–Stokes blow-up, Yang–Mills mass gap
cluster_10 — Algebraic geometry/motives	Birch and Swinnerton-Dyer conjecture	0.499	8	Hodge, Tate, standard conjectures, rational points, Langlands
unclustered	Quantum measurement problem	0.210	2	Quantum gravity, Hilbert's 6th problem

4. Minimum Collapsing Subset

A greedy set-cover over the strong-edge (≥0.5) transitive reach selects 14 problems that together cover all 67 unsolved problems in the artifact:

1. Algebrization barrier (p_np_algebrization_barrier)
2. Birch and Swinnerton-Dyer conjecture
3. Dark energy equation of state
4. Quantum measurement problem
5. Navier–Stokes existence and smoothness
6. abc conjecture
7. Singular Cardinal Hypothesis
8. Hilbert's 16th problem
9. Beal conjecture
10. Smooth 4D Poincaré conjecture
11. Generalized Poincaré conjecture (smooth category)
12. Borel conjecture
13. Cap set problem (exact growth)
14. Langlands program

If analogical / shared-technique links (≥0.3) are included, the greedy cover collapses to 4 problems: algebrization barrier, Schinzel's Hypothesis H, Borel conjecture, and cap set problem. This dramatic shrinkage shows how much of the artifact's connectivity is carried by cross-domain analogy rather than formal reduction.

5. Network Diagram

The diagram below shows the top leverage nodes and their strongest edges (weight ≥ 0.5). Thickness is omitted; all shown edges are high-confidence machinery/reduction links. Analogical weaker edges are suppressed to keep the diagram readable.

graph LR
    subgraph ComplexityCore [cluster_02 — Complexity core]
        PNP[p_vs_np]
        UGC[unique_games_conjecture]
        ETH[exponential_time_hypothesis]
        SETH[strong_exponential_time_hypothesis]
        BPP[bpp_vs_p]
        DLOG[discrete_log_in_p]
        FACT[factoring_in_p]
        GI[graph_isomorphism_in_p]
        NPI[np_intermediate_existence]
    end

    subgraph QuantumInfo [cluster_07 — Quantum & information]
        QS[quantum_supremacy_verification]
        YM[yang_mills_mass_gap]
    end

    subgraph PDE [cluster_09 — Fluid/field singularities]
        NS[navier_stokes_existence_smoothness]
        H16[hilbert_sixteenth_problem]
    end

    subgraph Foundations [cluster_06 — Logic/foundations]
        ALG[p_np_algebrization_barrier]
    end

    subgraph Arithmetic [cluster_04 — Arithmetic/Diophantine]
        TP[twin_prime_conjecture]
    end

    subgraph Motives [cluster_01/10 — L-functions & motives]
        RH[riemann_hypothesis]
        BSD[birch_swinnerton_dyer_conjecture]
        HODGE[hodge_conjecture]
        TATE[tate_conjecture]
    end

    PNP --> BPP
    PNP --> DLOG
    PNP --> ETH
    PNP --> FACT
    PNP --> GI
    PNP --> NPI
    PNP --> QS
    PNP --> SETH
    PNP --> UGC

    UGC --> PNP
    UGC --> ETH
    UGC --> SETH

    ETH --> SETH
    SETH --> ETH

    DLOG --> FACT
    FACT --> DLOG

    QS --> PNP
    QS --> YM

    YM --> H16
    NS --> H16
    H16 --> NS
    H16 --> YM

    ALG --> PNP

    RH --> TP
    TP --> RH

    BSD --> RH
    BSD --> HODGE
    BSD --> TATE
    HODGE --> TATE
    TATE --> HODGE

6. Caveats

• Analogy vs. rigorous reduction. Many cross-cluster edges in the artifact are 0.3–0.4 and are annotated as "shared techniques, cluster co-membership, and analogies." The strong-edge (≥0.5) analysis filters these out, but even the 0.5 threshold is a heuristic. Solving P vs NP will not automatically prove the cosmological constant problem or Hilbert's 6th problem; the artifact encodes a belief that progress on the complexity boundary propagates as methodology, not as formal implication.
• Cluster-internal vs. cross-cluster impact. The top global nodes are high-impact inside the complexity cluster and modestly bridge into quantum/PDE/foundations. Within arithmetic, the Riemann Hypothesis / abc / twin-prime triangle is far more levered than P vs NP, even though the global score is lower.
• Directionality. Edges represent "solving X unlocks Y," but reductions are not always one-way. Some pairs (P vs NP ↔ UGC, ETH ↔ SETH, factoring ↔ discrete log, abc ↔ Beal) are mutually linked, so either endpoint would collapse the other.
• Status bias. Several top-scoring nodes (bpp_vs_p, np_intermediate_existence, quantum_supremacy_verification, p_np_algebrization_barrier) are in HOLD status, meaning the artifact already considers them underspecified or blocked. High leverage does not imply high tractability.
• Coverage is not collapse. The set-cover result shows that 14 (or 4, with analogy) nodes touch every other problem in the graph. It does not mean proving those 14 would prove all 67; it means every other problem has at least one analogical or reduction path back to one of them.