Robust exponential lower bounds for fermionic and bosonic Gaussian ranks

arXiv:2610.02172 · quant-ph · Submitted 2026-10-01 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: Today's paper: "Robust exponential lower bounds for fermionic and bosonic Gaussian ranks".

Mira: Robust exponential lower bounds for Gaussian ranks in fermionic and bosonic systems establish fundamental limitations on classical simulation complexity,

Kai: First, who's behind it and why it matters.

Paper summary: Kai: So, to recap where we are, this paper "Robust exponential lower bounds for fermionic and bosonic Gaussian ranks" lays out a strong argument that non-Gaussianity forces an exponential increase in the complexity needed to decompose states into Gaussian ones.

Mira: Precisely; the thesis is that this decomposition complexity is inherently exponential in both fermionic and bosonic systems when you consider pure non-Gaussian states on finitely many modes.

Lev: It's interesting how they frame this as a problem related to coherent state rank, which connects it to a known area of study in quantum information theory.

Kai: They formally define Gaussian ranks, including the exact rank chi G(psi), the border rank chi G(psi), and the approximate rank chi((delta) G(psi).

Mira: The paper claims that for every pure non-Gaussian state on finitely many modes, there exists a constant such that the approximate border Gaussian rank of its tensor powers grows at least exponentially with the number of copies.

Lev: This means we're looking at an exponential lower bound on how many Gaussian states you need to approximate these target states when you take their tensor powers.

Kai: Specifically, they provide exponential lower bounds for the four-mode fermionic GHZ state and the bosonic single-photon state, using specific mathematical machinery.

Mira: The importance lies in showing that this complexity is robust; it holds at nonvanishing error and applies to a larger family of all pure Gaussian states including squeezed states.

Lev: That robustness is key for anyone trying to apply these bounds to real physical systems where perfect fidelity might not be achievable during the decomposition process.

Kai: So, in simple terms, the paper proves that if you have a non-Gaussian state in either system, simulating it by breaking it down into Gaussian parts gets exponentially harder as you scale up.

Mira: That's the big picture; it establishes a fundamental limitation on classical simulation complexity for these quantum systems due to their inherent non-Gaussian nature.

Lev: If this result is true, it suggests that any classical attempt to model these specific states will hit a computational wall related to this exponential growth.

Kai: So we've covered the main ideas; now let's look at what this all means for the broader landscape of quantum simulation.

Conclusion: Mira: Wrapping up our discussion on "Robust exponential lower bounds for fermionic and bosonic Gaussian ranks," the authors have provided robust mathematical proofs showing that non-Gaussianity universally leads to an exponential increase in the decomposition complexity of these states.

Kai: I think the real significance is that they're setting a hard mathematical boundary on classical simulation complexity, demonstrating precisely how difficult it is to decompose these states into Gaussian components.

Lev: For quantum error correction researchers like myself, this result gives us a benchmark for understanding the resources required to represent non-Gaussian states in terms of Gaussian ones.

Kai: It shows that for certain physical systems, you cannot rely on simple polynomial scaling for simulation complexity when dealing with these specific states under these conditions.

Mira: The authors' work is important because it defines this universal exponential relationship, which applies across both fermionic and bosonic systems without needing to specify the state's energy.

Lev: If we can use these bounds, it provides a rigorous justification for why certain computational tasks involving non-Gaussian states are inherently limited in classical decomposition methods.

Kai: It’s about understanding the intrinsic difficulty of working with these quantum states from a simulation standpoint.

Fuchuan Wei, *Kong-Wing Wu*, *Zhengwei Liu*, +Zi-Wen Liu

Yau Mathematical Sciences Center, Tsinghua University · Qiuzhen College, Tsinghua University · Department of Mathematics, Tsinghua University · Yanqi Lake Beijing Institute of Mathematical Sciences and Applications

quant-ph

Submitted: 2026-10-01

Updated: 2026-10-01

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 90/100

The gist: Robust exponential lower bounds for Gaussian ranks in fermionic and bosonic systems establish fundamental limitations on classical simulation complexity, proving that non-Gaussianity universally

Key concepts

Exact Gaussian Rank ($\chi_G(|\psi\rangle)$)
This is the smallest number of pure Gaussian states needed to form a linear combination that exactly equals the target state $|Κ\rangle$. It measures the inherent complexity of representing a state using only Gaussian components.
Border Gaussian Rank ($\chi_G(|\psi\rangle)$)
This is the smallest integer $r$ such that the target state can be approximated arbitrarily closely by vectors composed of Gaussian states with rank at most $r$. If no finite $r$ works, the rank is considered infinite.
Approximate Gaussian Rank ($\chi(\delta)_G(|\psi\rangle)$)
This measures how close the target state $|Κ\rangle$ is to a vector $|Λ\rangle$ that has a low Gaussian rank. It is defined as the minimum distance to such an approximant, quantifying the difficulty of finding a simple Gaussian representation.
Non-Gaussianity
This refers to quantum states that cannot be described solely by Gaussian distributions. The paper shows that even small amounts of non-Gaussian behavior force the required Gaussian decomposition complexity to grow exponentially with the number of copies.

Terminology

Summary

Robust exponential lower bounds for Gaussian ranks in fermionic and bosonic systems establish fundamental limitations on classical simulation complexity, proving that non-Gaussianity universally entails exponential Gaussian decomposition complexity for both fermionic and bosonic systems.

The gist

Non-Gaussianity alone forces exponential rank growth: Theorem 1 states that for a fixed pure non-Gaussian state on finitely many modes, there exists a constant such that the approximate border Gaussian rank of its tensor powers grows at least exponentially with the number of copies, specifically showing an exponential lower bound for the four-mode fermionic GHZ state and the bosonic single-photon state.

Defining Gaussian Ranks

The paper formally defines several measures related to Gaussian states. The exact Gaussian rank, denoted as chiG(sigma⟩), is the smallest number of pure Gaussian states needed to express a target state sigma⟩ as a linear combination. The border Gaussian rank, chiG(sigma⟩), is the smallest integer for which sigma⟩ can be approximated arbitrarily closely in norm by vectors of Gaussian rank at most r; if no finite r suffices, the rank is +∞. The approximate Gaussian rank, chi(delta)G(sigma⟩), is defined as the minimum distance to a normalized approximant alpha⟩ in norm: chi((δ)G(sigma⟩):= min∥ sigma⟩− alpha⟩∥2≤ δ chiG(alpha⟩).

Fermionic Lower Bounds

The paper establishes exponential lower bounds for fermionic systems. Theorem 2 proves that there exist parameters such that for every fixed 0 ≤ delta < 1 and all sufficiently large k, the approximate border Gaussian rank of psi⟩⊗k grows at least exponentially: chi((δ)G(sigma⟩⊗k) ≥ e c k. This is achieved by reducing the target state to four modes and using a lower bound on Gaussian rank there, transferring it back with controlled approximation error. The explicit exponential base for the four-mode fermionic GHZ state is derived as 343/243.

Bosonic Lower Bounds

For bosonic systems, Theorem 4 provides an exponential lower bound for any fixed pure non-Gaussian state on finitely many modes, without any assumption on its energy. The result holds for the bosonic single-photon state 1⟩⊗k, yielding a lower bound of (11/10)k at zero error. The proof involves reducing the target by Gaussian postselection to states near a nonvacuum Fock state and establishing an exponential rank bound for their products, which is then transferred back with controlled approximation error.

Mechanism of the Lower Bound

The core mechanism relies on combining several techniques:

  1. Reduction to four modes: The target state is reduced to four modes using a Gaussian unitary and postselection, resulting in a state Mt⟩ (for fermions) or 1⟩⊗k (for bosons).

  2. Matrix Rank Bounds: For the fermionic case, the lower bound is derived by comparing the rank of a matrix constructed from Majorana monomials on the target state with its rank on an approximant state. This involves analyzing subadditivity of matrix rank and using bounds derived from singular value decomposition (Lemma 15).

  3. Entropy Maximization: For both systems, Theorem 20 utilizes a saddle-point estimate for coefficients of large powers, where the base B t is determined by maximizing an entropy functional over three real parameters. This yields a continuous exponential growth rate independent of the error tolerance.

Key Results and Dichotomy

The results show that non-Gaussianity universally entails exponential Gaussian decomposition complexity. For four-mode fermionic states, the exact border rank is 3, while the approximate border rank is bounded below by (1 - δ 2) D k,p C4k,p. For the bosonic single-photon state 1⟩⊗k, the exact border rank is +∞ while its approximate border rank grows exponentially at a base of 11/10. This dichotomy demonstrates that Gaussian states satisfy chi((δ)G(sigma⟩⊗k) = 1, highlighting the fundamental limitations of decomposition-based classical simulation for these systems.

Conclusion

The paper proves that for every fixed pure non-Gaussian state on finitely many modes, there are positive integers m k with m k/k → p psi such that the approximate border Gaussian rank grows exponentially at a rate proportional to p psi log B t.

Improvements for AI systems

This paper establishes fundamental limitations on the classical simulation complexity of quantum states, specifically proving robust exponential lower bounds for Gaussian ranks in both fermionic and bosonic systems.

Here are the specific improvements to AI systems that can be derived from this research:

  1. Improve the efficiency of quantum state decomposition and simulation algorithms by establishing hard limits on classical resources required for decomposition.

  2. Develop more efficient free-state decompositions for quantum computation by leveraging the known Gaussian structure, which is classically tractable but non-Gaussian states are not.

  3. Design new classical simulation complexity metrics that provide unconditional exponential lower bounds rather than subexponential or quadratic ones for simulating tensor powers of non-Gaussian states.

Specific capabilities of the improved AI system:

  1. The AI system can rigorously determine the minimum number of Gaussian components (the Gaussian rank) required to represent a target quantum state, providing guaranteed exponential complexity scaling with the number of copies, even when approximation errors are considered.

  2. The system can predict and quantify the exponential growth rate of simulation complexity for non-Gaussian quantum systems (both bosonic and fermionic), setting fundamental theoretical boundaries on how efficiently classical computers can simulate these states.

  3. The AI system can optimize quantum circuit synthesis or state preparation protocols by identifying inputs that are inherently difficult to simulate classically, thereby guiding researchers toward more efficient quantum-native computational approaches rather than relying solely on classical decomposition methods.

  4. The system can serve as a diagnostic tool for quantum hardware simulators, flagging when a simulation complexity bottleneck is caused by the inherent non-Gaussian nature of the state being simulated, allowing for targeted optimization of the simulation method rather than generic resource scaling.

Sources

Related papers