← Domain Convergence Research

Non-Abelian Anyon Family Sweep

Which anyon fusion algebrasRules governing what happens when two anyons combine — a multiplication table for particle types. The algebra determines which output particles are possible from each input pair. support complex systems via self-catalysis, RAF closure, and computational universality?

Summary

Family Generator Self-catalytic? RAF closure? Universal braiding? Dim. growth Fusion outcomes Suitability
SU(2)2 / Ising σ No (2-step cycle) Partial No √2 ≈ 1.41 2 Low
SU(2)3 / Fibonacci τ Yes Yes Yes φ ≈ 1.62 2 High
SU(2)4 (j=1) j=1 Yes Yes Not by braiding alone 2.00 3 High
SU(2)5 / SO(3)5 j=1 Yes Yes Yes ≈2.25 3 High
SU(2)6 (j=1) j=1 Yes Yes Yes ≈2.41 3+ High
SU(3)2 fund. Yes Yes Likely ≈1.93 2–3 Medium-High
SO(3)2 / Metaplectic φ Context-dep. Partial No ≈1.41 2 Low
D(S3) eC Yes Yes No 2.00 3 Medium
Haagerup α Yes Yes Open question ≈4.30 4+ Uncertain (exotic)

Detailed Analysis

1. SU(2)k Chern-Simons — The Main Sequence

The most important family. Anyons labeled by spin j = 0, 1/2, 1, ..., k/2. Fusion follows truncated Clebsch-Gordan rulesStandard rules for combining angular momenta in quantum mechanics. In anyon physics, the truncated version (cut off at level k) determines which particle types can result from fusing two anyons.: j1 × j2 = |j1−j2| + ... + min(j1+j2, k−j1−j2). Quantum dimensionA number measuring how fast the computational state space grows when adding particles. Not spatial dimensions. Fibonacci anyons have quantum dimension φ ≈ 1.618 — the golden ratio.: dj = sin(π(2j+1)/(k+2)) / sin(π/(k+2)).

k=2 (Ising anyons) — {1, σ, ψ}

Fusion rules:

σ × σ = 1 + ψ
σ × ψ = σ
ψ × ψ = 1

Self-catalysis: σ×σ does NOT produce σ. However, σ×σ→ψ then σ×ψ→σ gives a two-step regeneration cycle. This is catalytic cycling but not direct self-catalysis.

RAF closure: Partial. σ generates {1,ψ} directly but needs ψ as co-reactant to regenerate itself. The full set {1,σ,ψ} is closed under fusion. The RAF condition is technically met (σ catalyzes ψ production, ψ catalyzes σ regeneration), but the catalytic network is a simple 2-cycle, not a complex web.

Universality: No. Ising braiding generates only the Clifford groupA finite set of quantum gates that can efficiently simulate stabilizer circuits but cannot perform universal quantum computation alone. The "safe" subset — powerful but limited. — not dense in SU(N). Requires magic state distillationA technique to boost limited gates to universality by preparing special quantum states. Expensive — requires many noisy copies to produce one clean state. Unnecessary when braiding alone is universal. for universality.

Dimension growth: dσ = √2 ≈ 1.414. Slowest growth of any non-abelian family.

Too simple. The fusion algebra is a 2-cycle, not a complex network. The failure of braiding universality correlates with the failure of direct self-catalysis.

k=3 (Fibonacci anyons) — {1, τ}

(SO(3)3 subtheory, equivalently the integer-spin sector)

Fusion rules:

τ × τ = 1 + τ

Self-catalysis: Yes. τ produces itself directly.

RAF closure: Yes. All three Kauffman criteria satisfied: (1) Generation — τ generates {1, τ} from a single fusion. (2) Transformation — fusion transforms particle types. (3) Closure — all outputs ∈ {1, τ}.

Universality: Yes. Fibonacci anyon braiding is dense in SU(N)The braiding operations can approximate ANY quantum gate to arbitrary precision. Like how rational numbers are dense in the reals — there's always one close enough to any target value. — computationally universal.1

Dimension growth: dτ = φ = (1+√5)/2 ≈ 1.618. The golden ratio. Fusion space dimensions follow the Fibonacci sequence: 1, 1, 2, 3, 5, 8, 13, ...

The minimal universal model. Simplest possible self-catalytic, RAF-closed, computationally universal fusion algebra. The "sweet spot."

k=4 — {1, j=1/2, j=1, j=3/2, j=2}

Fusion rules for the generating anyon j=1:

1 × 1 = 0 + 1 + 2  (three outcomes)
1 × 1/2 = 1/2 + 3/2
1 × 3/2 = 1/2 + 3/2
1 × 2 = 1
3/2 × 3/2 = 0 + 1
2 × 2 = 0

Self-catalysis: Yes. j=1 fusing with itself produces j=1.

RAF closure: Yes. j=1 generates the full anyon set through iterated fusion. All elements transform others. Closure holds (truncation ensures all outputs are valid anyons).

Universality: Not by braiding alone. SU(2)4 is one of the two known exceptions (with k=8) to the Freedman–Larsen–Wang density theorem — the braid-group image is not dense in SU(N).1 Reaching universality therefore requires supplementary operations (measurement, or ancillary abelian charges), not braiding on its own.

Dimension growth: d1 = sin(3π/6)/sin(π/6) = 2.000. Integer quantum dimension. Fusion space grows as 2n.

Self-catalytic with richer fusion (3 outcomes vs 2), though — unlike k=3, 5, 6 — not universal by braiding alone. The integer quantum dimension also means less "irrational complexity" in the phase structure. The three-outcome fusion still creates a richer catalytic network than τ×τ = 1+τ.

k=5 (SO(3)5 integer-spin sector) — {0, 1, 2}

Fusion rules:

1 × 1 = 0 + 1 + 2
1 × 2 = 1 + 2
2 × 2 = 0 + 1

Self-catalysis: Partial. j=1 is self-catalytic (1×1 = 0+1+2 contains 1); j=2 is not self-producing (2×2 = 0+1 does not contain 2), though it is generated by the network (1×1 and 1×2 both yield 2).

RAF closure: Yes — every non-trivial object is generated from the network on the food set {0}. But it is not fully connected: j=2 does not catalyze its own production, so this is not a maximally-autocatalytic web.

Universality: Yes. Braiding is dense in SU(N).

Dimension growth: d1 = 2cos(2π/7) + 1 ≈ 2.247. Irrational, like Fibonacci but with a different algebraic number.

RAF-closed with a self-catalytic j=1 generator, but not the fully-connected web the fusion table first suggests — j=2 is produced by the network yet is not self-catalytic. Still a richer catalytic network than Fibonacci; the tradeoff is that more particle types mean more complex experimental realization.

k=6 and beyond

The pattern continues: more anyon types, more fusion channels, richer catalytic networks. Self-catalysis and RAF closure are generic for k≥3. Braiding universality holds for all k≥3 except k=4 and k=8, where the braid-group image is not dense and braiding must be supplemented by additional operations.

Key pattern in SU(2)k:

  • k=2: No self-catalysis, no universality, simple 2-cycle. Below threshold.
  • k=3: Minimal self-catalysis, universal, RAF closed. At threshold (the "Fibonacci edge").
  • k≥4: Rich self-catalysis, strongly RAF closed, universal by braiding alone for every k except k=4 and k=8. Above threshold.

The onset at k = 3

Both panels run the same experiment: seed one non-abelian anyon, fuse it with itself, and follow where the products go. Watch whether the seed type ever reappears as a direct product of its own fusion.

k = 2 — Ising, {1, σ, ψ}
σ × σ = 1 + ψ  ·  σ × ψ = σ
σ × σ does not contain σ.
k = 3 — Fibonacci, {1, τ}
τ × τ = 1 + τ
τ × τ contains τ.

Press Play. Nothing has been fused yet.

There is a phase transition at k=3 — the onset of both self-catalysis and computational universality simultaneously. This is not coincidental: the same algebraic structure (non-trivial self-coupling in the fusion rules) enables both properties.

2. SU(3)k — Richer Gauge Group

SU(3) theories have more anyon types at each level (labeled by Young tableaux).

SU(3)1

Abelian. Three anyons forming Z3. No interest for complex systems.

SU(3)2

Six anyon types. The fundamental representation has quantum dimension d ≈ 1.932 (irrational, related to tribonacci-like algebraic numbers).

Fusion: f×f = f̅ + (adjoint sector). Self-catalysis depends on whether the adjoint includes f. In SU(3)2, the truncation makes this nontrivial: f×f̅ = 1 + adj, and adj×f = f + f̅ + .... The catalytic network is richer than SU(2) at comparable levels because of additional structure from the larger gauge group.

Likely self-catalytic and RAF closed, with richer catalytic networks than SU(2) at comparable k. Braiding likely universal. The tradeoff: more particle types makes physical realization harder. Confidence in the specific fusion rules is lower here — SU(3) Chern-Simons theory is well-defined but less commonly tabulated in full detail than SU(2).
3. SO(3)k / Metaplectic Anyons

SO(3)k has only integer-spin anyons (j=0,1,...,k). This is not the same as restricting SU(2)2k to integer spins (the relationship involves different modular data).

SO(3)2 (Metaplectic): {1, s, φ}

φ × φ = 1 + s
φ × s = φ
s × s = 1

Structurally identical to Ising (same fusion rules with relabeling σ→φ, ψ→s).

Self-catalysis: No (same as Ising — 2-step cycle only). RAF closure: Partial. Universality: No.

SO(3)k for k≥3

Recovers the integer-spin sector patterns from SU(2)2k. At k=3, the SO(3)3 theory gives the Fibonacci anyon model {1,τ}. The "good" properties emerge here too at k≥3.

The metaplectic label specifically refers to SO(3)2, which shares Ising's limitations. The SO(3)k family as a whole follows the same phase transition at k=3.
4. D(S3) — Quantum DoubleA mathematical construction that builds an anyon theory from any finite group. The resulting anyons come from the group's conjugacy classes and representations. Produces non-abelian anyons but with limited (non-universal) braiding. of S3

The simplest non-abelian finite group quantum double. 8 anyon types corresponding to irreps of the Drinfeld doubleThe algebraic structure underlying quantum doubles. Named after mathematician Vladimir Drinfeld. Combines a group with its dual to create a quasi-triangular Hopf algebra — the mathematical home of anyon fusion and braiding. of the symmetric group S3. Non-abelian anyons include a 2-dimensional charge sector (eC, associated with the conjugacy class of transpositions).

Fusion rules for eC:

eC × eC = 1 + eC + eC′

Self-catalysis: Yes. eC × eC produces eC.

RAF closure: Yes. The generating anyon produces multiple types including itself.

Universality: No. D(S3) braiding is NOT universal — it generates only a finite group of gates. This is a general property of quantum doubles of finite groups: the braiding is too "rigid."

Dimension growth: deC = 2. Integer quantum dimension. Growth is 2n.

Self-catalytic and RAF-closed, but NOT computationally universal — its braiding generates only a finite gate group. This breaks the apparent correlation between RAF closure and universality seen in SU(2)k. Nor is there a classical loophole: group multiplication over S3 does not give universal classical computation either, because S3 is solvableA group built by extension from abelian pieces. By Barrington's theorem, bounded-width computation over a group is universal only for a non-solvable group — the smallest is A₅ (order 60; Barrington used S₅) — so S₃ multiplication is not a route to classical universality. (Barrington's theorem needs a non-solvable group; the smallest is A5 ⊆ S5). The break with universality here is therefore genuine.
5. Haagerup Fusion Category

The Haagerup category is exotic — it does not arise from any Chern-Simons theory or quantum group construction. Discovered as a subfactorA mathematical structure from operator algebra — a way of embedding one algebra inside another. Some fusion categories, like Haagerup, were discovered as subfactors rather than from any physical gauge theory., it has 6 simple objects.

The generating object α has quantum dimension dα = (3+√13)/2 ≈ 3.30 (its subfactor index is the larger (5+√13)/2 ≈ 4.30) — among the largest for any known "small" MTCModular Tensor Category — the mathematical framework for classifying anyon theories. A complete MTC specifies all particle types, fusion rules, braiding matrices, and topological invariants of a 2+1D topological phase..

Fusion rules (partial; full rules involve 6 objects):

α × α = 1 + α + α* + β + β* + γ  (produces essentially everything)
α × α* = 1 + α + α* + ...

Self-catalysis: Yes, emphatically. α×α produces α.

RAF closure: Yes, strongly. A single fusion event generates most of the anyon set.

Universality: Open question. Whether Haagerup braiding is computationally universal is not definitively settled in the literature. The large quantum dimension suggests rich braiding, but universality proofs require detailed analysis of the braid representations.

Dimension growth: dα ≈ 4.303. Extremely fast — fusion space grows as ~4.3n.

Maximally self-catalytic (one fusion produces nearly everything), but the fusion algebra may be too rich — like a chemical system where every reaction produces every product. For complex systems, you need structure in the catalytic network, not just maximal connectivity. A fully connected network is actually less interesting than one with intermediate connectivity (same principle as K≈2 vs K=N in Boolean networks).

Patterns

  1. Self-catalysis is generic, not special. Every non-abelian anyon family with k≥3 (or equivalent complexity threshold) exhibits self-catalysis. The only exception is the Ising/metaplectic family (k=2), which has a 2-step catalytic cycle instead. Implication for the AC bridgeThe proposed correspondence between Topology (A) and Autocatalysis / Closure (C) — one of the cross-domain bridges this research program is testing, not an established result.: the proposed correspondence between anyon fusion and RAF closure is not unique to Fibonacci anyons. It appears to be a generic property of sufficiently complex fusion algebras.
  2. There is a phase transition at the onset of self-catalysis. In SU(2)k: k=2 has no self-catalysis and no universality; k=3 has both. The transition is sharp. The same algebraic structure (non-trivial self-coupling Njjj ≥ 1) enables both self-catalysis and computational universality simultaneously.2 This is the key finding: the two properties share a common algebraic origin.◆ our resultAn original, not-yet-peer-reviewed result from this research program — distinct from established consensus.
  3. The Fibonacci model is the critical case, not the typical case. Fibonacci sits exactly at the phase transition: minimal self-catalysis (one outcome: τ×τ→τ plus vacuum), minimal universality (barely dense in SU(2)), golden ratio dimension growth (the "slowest irrational"). Analogous to K=2 in Boolean networks — the critical point between order and chaos. The Fibonacci model is the edge of chaosThe boundary between ordered and chaotic behavior where computation is maximized. Systems here are complex enough to process information but ordered enough to maintain structure. Fibonacci anyons sit at this boundary for fusion algebras. for anyon fusion algebras.
  4. Universality does not perfectly correlate with RAF closure. D(S3) is self-catalytic and RAF-closed but not computationally universal — its braiding yields only a finite gate group. The correlation holds within SU(2)k but genuinely breaks for D(S3): because S3 is solvableA group built by extension from abelian pieces. By Barrington's theorem, bounded-width computation over a group is universal only for a non-solvable group (the smallest is A₅ ⊆ S₅), so S₃ multiplication is not a classical-universality loophole., group multiplication over it does not give universal classical computation either (Barrington's theorem needs a non-solvable group; A5 ⊆ S5 is the smallest). The correlation is not rescued by a classical PCE loophole here.
  5. There is a "sweet spot" in fusion algebra complexity.
    RegimeExampleSelf-cat.UniversalNetworkComplex?
    Too simpleIsing (k=2)NoNo2-cycleNo
    CriticalFibonacci (k=3)Yes (minimal)Yes (barely)Self-loop + vacuumEdge
    ComplexSU(2)4–5Yes (rich)YesMulti-loop webRich
    SaturatedHaagerupYes (maximal)OpenFully connectedToo uniform?
    The sweet spot is k=3 through k≈5–6. Below k=3, the algebra is too simple. Much above k=6, the fusion algebra becomes so richly connected that it may lose the structured heterogeneity needed for complex dynamics.
  6. The golden ratio is not unique, but it is special. φ is the smallest quadratic irrational with a self-similar continued fraction expansion [1;1,1,1,...]. This connects to the Fibonacci sequence, which is itself the simplest non-trivial recursive sequence. The "minimal complexity" of φ mirrors the "minimal self-catalysis" of the Fibonacci anyon model. Other growth rates (SU(2)5: 2cos(2π/7)+1, Haagerup: (5+√13)/2) involve different algebraic numbers. Whether these differences matter for complex system dynamics is an open question.

Conclusions for the AC Bridge

  1. The AC bridge via RAF closure is generic across anyon families with k≥3. It is not a special property of Fibonacci anyons. Any sufficiently complex fusion algebra satisfies autocatalytic closure.3
  2. Fibonacci is the critical/minimal case — the phase transition between algebras that do and don't support complex systems. This makes it the natural starting point for investigation, but not the only target.
  3. The k=4 and k=5 models may be better targets for finding complex dynamics in lattice simulations, because their richer fusion rules (3+ outcomes) create denser catalytic networks with more room for structured heterogeneity.
  4. The onset of self-catalysis coincides with the onset of computational universality (at k=3). We read this as suggestive evidence for the proposed AC-bridge correspondence◆ our resultAn original, not-yet-peer-reviewed reading from this research program — a proposed correspondence, not established consensus.: the same algebraic structure enables both the topological property (universal computation) and the autocatalytic property (self-sustaining closure). Whether they truly share a common algebraic root, rather than coinciding, remains to be established — and the k=4 / k=8 braiding exceptions show the coincidence is not perfect.
  5. Open questions: Does the lattice braiding search find Class IV behavior more easily at k=4–5 than k=3? Is the "sweet spot" at k=3 (edge of chaos) or k=4–5 (richly complex)? Does the Haagerup category produce complex dynamics despite its non-standard origin? Is classical computational equivalence sufficient for the bridge, or is quantum universality required?
  1. Freedman, M. H., Larsen, M. & Wang, Z. (2002). A modular functor which is universal for quantum computation. Communications in Mathematical Physics, 227, 605-622. Establishes that SU(2)k braiding is dense in SU(N) for k ≥ 3, with the two exceptions k = 4 and k = 8, where the braid-group image is finite and braiding alone is not universal. ↑
  2. Kitaev, A. Y. (2003). Fault-tolerant quantum computation by anyons. Annals of Physics, 303(1), 2-30. Foundational paper establishing topological quantum computation via non-abelian anyons. ↑
  3. Hordijk, W. & Steel, M. (2004). Detecting autocatalytic, self-sustaining sets in chemical reaction systems. Journal of Theoretical Biology, 227(4), 451-461. Formal definition and detection algorithm for RAF sets. ↑
← Domain Convergence Research · Web Annex