Violation of the Universal Operator Growth Hypothesis in W3 Conformal Field Theories ✛ ✎
Context
The study of operator growth in quantum systems relies on the universal operator growth hypothesis, which states that Lanczos coefficients grow at most linearly. Most tests have been performed in models with Virasoro symmetry or on lattices. Extended algebras, such as W-algebras, arise in the physics of the fractional quantum Hall effect, critical statistical systems, and SYK-like models, but their influence on operator growth remained unexplored. This work fills that gap by considering CFTs with W3 symmetry.
Methods
The authors consider a W3 conformal field theory with large central charge, using a free-field realization with two scalar fields. The Liouvillian is chosen as L = κ1(L1 + L−1) + κ2(W2 + W−2) to directly involve the spin-3 generator. The Lanczos coefficients are computed in the module of a heavy primary state by acting with W−2 on an arbitrary descendant state expressed in terms of oscillator modes. Nineteen classes of coefficients are obtained, ten of which exhibit superlinear growth.
Results
Two classes of coefficients, denoted bI and bII, grow asymptotically as $N^2$, where N is the descendant level. This is the maximal violation of the linear bound. Other classes grow as $N^{3/2}$, $N^{5/4}$, and so on. For typical descendants in the large-N limit, the growth of coefficients is $N^{1/4}$, $N^{1/2}$, and $N^{3/4}$; the last is new compared to Liouville theory. The quadratic growth is also reproduced in the global SL(3,R) subalgebra, indicating that the root of the violation lies in extended higher-rank symmetry rather than in the infinite-dimensionality or nonlinearity of the full W3 algebra.
Implications
The results show that extended W-symmetries qualitatively change operator growth, allowing known bounds on information scrambling speed to be circumvented. This calls into question the universality of the operator growth hypothesis in theories with higher spins. The work also suggests that for $W_n$ symmetry, the maximal growth of Lanczos coefficients may scale as $N^{n-1}$.
Future development
A natural next step is to test the hypothesis of $N^{n-1}$ growth for $W_n$ algebras. It is also interesting to explore connections with generalized DSSYK models and four-dimensional gauge theories via the AGT and Gaiotto–Verlinde correspondences. One could also study the impact of the violation on thermalization and scrambling in systems with higher spins.
Impact
The results may affect our understanding of quantum chaos in systems with extended symmetry, including the fractional quantum Hall effect and critical statistical models. They also open the way to revisiting the bounds on Krylov complexity in conformal field theories.
Next steps
The next step could be to compute the Lanczos coefficients for W_n algebras and verify the proposed scaling. It is also important to investigate how the violation manifests in physical observables such as correlation functions or scrambling rates.
Key open problems
This work touches on the fundamental question of the limits of quantum complexity growth and the connection between symmetry and chaos. It shows that extended symmetries can lead to superexponential growth of complexity, challenging conventional views on thermalization in quantum systems.
The knowledge machine's view
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
This paper examines the growth of operators in W3 conformal field theories. The authors compute Lanczos coefficients in the module of a heavy primary state using a Liouvillian with a spin-3 generator, and find classes of coefficients with asymptotic growth up to N^2, which violates the universal linear bound. They show that the violation already occurs in the global subalgebra SL(3,R), pointing to a connection with extended symmetry.
Where its strength is
The strength of the work lies in the explicit computation of Lanczos coefficients for the W3 algebra and the demonstration of superlinear growth, including quadratic growth, with the identification of the source of the violation in the global subalgebra.
Where it could go next
- It might be interesting to see how the abrupt growth described in a neighboring work on ice patterns relates to the discovered classes of Lanczos coefficients and their asymptotics.based on: Frosted Pattern: When Growth Leaps Forward arXiv:2607.01142
- Nearby there is work on complexity on graphs with quantum walks; perhaps one should consider the analogy between time acceleration in such systems and superlinear growth of operators in W3 theories.based on: Complexity on Graphs: How Quantum Walks Speed Up Time arXiv:2602.04949
- In the archive there is an analysis of a torus universe with a quantum jackpot; perhaps methods or ideas from there could help investigate the influence of topology or additional symmetries on operator growth.based on: Torus Universe: A Quantum Jackpot for the Big Bang arXiv:2605.05317v2
Worth trying
- Test whether the quadratic growth reproduces for other W-algebras (e.g., W4) and whether it agrees with the N^{n-1} hypothesis.
- Investigate how abrupt changes in the growth of Lanczos coefficients relate to discrete structures like ice patterns.
- Consider whether superlinear growth can be interpreted as accelerated information propagation, analogous to effects in quantum walks on graphs.
Nearby works in our archive
- Frosted Pattern: When Growth Leaps Forwardsimilarity 0.50
- Complexity on Graphs: How Quantum Walks Speed Up Timesimilarity 0.50
- The Universe's Size and Expansion Are Uncertainsimilarity 0.48
- Torus Universe: A Quantum Jackpot for the Big Bangsimilarity 0.47
- How to Keep a Problem from Ballooning: A Fast Method for Machine Learningsimilarity 0.47
- Every Ninth Zero: A Mathematical Sieve on the River of Primessimilarity 0.47
The archive is sparse around this work: the closest review is only 0.50, against a usual 0.64. It looks like few have drilled nearby.
The Lanczos coefficients are named after Cornelius Lanczos, who developed an algorithm for solving systems of linear equations in the 1950s, unaware of its future application in quantum complexity theory.
The idea that extended symmetry can accelerate complexity growth echoes the concept of the 'bureau of chaos' in Neal Stephenson's novel 'Anathem', where mathematical structures influence physical processes.
Key numbers
- Maximal growth of Lanczos coefficients: N^2
- Number of coefficient classes: 19
- Number of classes with superlinear growth: 10
- Growth for typical descendants: N^{1/4}, N^{1/2}, N^{3/4}