How hard is it to approximate the Jones polynomial?

Freedman, Kitaev, and Wang [arXiv:quant-ph/0001071], and later Aharonov, Jones, and Landau [arXiv:quant-ph/0511096], established a quantum algorithm to “additively” approximate the Jones polynomial V(L,t) at any principal root of unity t. The strength of this additive approximation depends exponentially on the bridge number of the link presentation. Freedman, Larsen, and Wang [arXiv:math/0103200] established that the approximation is universal for quantum computation at a non-lattice, principal root of unity; and Aharonov and Arad [arXiv:quant-ph/0605181] established a uniform version of this result. In this article, we show that any value-dependent approximation of the Jones polynomial at these non-lattice roots of unity is #P-hard. If given the power to decide whether |V(L,t)| > a or |V(L,t)| < b for fixed constants a > b > 0, there is a polynomial-time algorithm to exactly count the solutions to arbitrary combinatorial equations. In our argument, the result follows fairly directly from the universality result and Aaronson’s theorem that PostBQP = PP [arXiv:quant-ph/0412187].

💡 Research Summary

The paper investigates the computational hardness of approximating the Jones polynomial V(L, t) at principal roots of unity that are not lattice points. The starting point is the quantum algorithm introduced by Freedman, Kitaev, and Wang (2000) and later refined by Aharonov, Jones, and Landau (2005), which provides an additive approximation of V(L, t). The error of this approximation scales exponentially with the bridge number of the link presentation, meaning that only links with small bridge number can be approximated efficiently. Freedman, Larsen, and Wang (2001) showed that for non‑lattice principal roots the same algorithm is universal for quantum computation: any quantum circuit can be encoded into a link such that the value of the Jones polynomial at the chosen root encodes the circuit’s output probability. Aharonov and Arad (2006) strengthened this result by giving a uniform version that works for all input sizes with the same polynomial overhead.

The authors of the present work ask a more refined question: suppose we are given two fixed constants a > b > 0 and a decision oracle that, for any link L, tells us whether |V(L, t)| > a or |V(L, t)| < b (the “gap‑decision” problem). They prove that such an oracle would allow us to solve any #P‑complete counting problem in polynomial time, i.e., the gap‑decision problem is #P‑hard. The proof proceeds by directly linking the universality construction to Aaronson’s theorem that PostBQP = PP.

The reduction works as follows. Take an arbitrary Boolean circuit C whose number of satisfying assignments we wish to count. Using the standard topological quantum field theory (TQFT) encoding, construct a link L(C) whose Jones polynomial at the chosen non‑lattice root t satisfies
 |V(L(C), t)| = α·|#(C = 1) − #(C = 0)|,
for some non‑zero complex constant α that depends only on t. Consequently, if |V(L(C), t)| exceeds a, the circuit has many 1‑outputs; if it falls below b, the circuit has few. By binary searching over the possible numbers of satisfying assignments and invoking the gap‑decision oracle at each step, we can determine the exact count of satisfying assignments. Because counting solutions to Boolean formulas is #P‑complete, the existence of a polynomial‑time algorithm for the gap‑decision problem would imply #P ⊆ P, collapsing the polynomial hierarchy. This contradicts widely believed complexity‑theoretic assumptions, establishing #P‑hardness.

The argument crucially uses Aaronson’s PostBQP = PP result: the ability to post‑select on measurement outcomes in a quantum computation elevates the computational power to PP, which captures #P counting. The universality of the Jones polynomial at non‑lattice roots already provides a way to embed any quantum circuit into a link; adding post‑selection translates the gap‑decision oracle into a PP oracle, completing the reduction.

Beyond the hardness result, the paper discusses implications for quantum algorithms. The known additive approximation algorithm’s dependence on bridge number appears unavoidable: any improvement that removed this exponential dependence would, by the authors’ argument, give a polynomial‑time algorithm for a #P‑hard problem. Thus the current state of the art reflects a genuine complexity barrier rather than a technical limitation.

The authors conclude with several open directions. First, whether similar hardness holds for lattice roots of unity (where the underlying anyonic models are not universal) remains open. Second, the possibility of alternative approximation schemes that avoid bridge‑number dependence but still provide useful information about V(L, t) is an intriguing challenge. Finally, exploring models that capture the power of post‑selection without invoking unphysical assumptions could shed light on the precise boundary between efficiently simulable quantum systems and those that are computationally universal.

In summary, the paper establishes that any value‑dependent approximation of the Jones polynomial at non‑lattice principal roots of unity is #P‑hard. The result follows almost immediately from the known universality of the Jones polynomial for quantum computation and Aaronson’s PostBQP = PP theorem, thereby linking topological invariants, quantum algorithms, and classical counting complexity in a strikingly tight way.