Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:1101.5355 · Electronic colloquium on computational complexity

Advice Coins for Classical and Quantum Computation

Scott Aaronson, Andrew Drucker

Published Jan 27, 2011• 1 citation
arXivHTMLPDF

Abstract

We study the power of classical and quantum algorithms equipped with nonuniform advice, in the form of a coin whose bias encodes useful information. This question takes on particular importance in the quantum case, due to a surprising result that we prove: a quantum finite automaton with just two states can be sensitive to arbitrarily small changes in a coin's bias. This contrasts with classical probabilistic finite automata, whose sensitivity to changes in a coin's bias is bounded by a classic 1970 result of Hellman and Cover. Despite this finding, we are able to bound the power of advice coins for space-bounded classical and quantum computation. We define the classes BPPSPACE/coin and BQPSPACE/coin, of languages decidable by classical and quantum polynomial-space machines with advice coins. Our main theorem is that both classes coincide with PSPACE/poly. Proving this result turns out to require substantial machinery. We use an algorithm due to Neff for finding roots of polynomials in NC; a result from algebraic geometry that lower-bounds the separation of a polynomial's roots; and a result on fixed-points of superoperators due to Aaronson and Watrous, originally proved in the context of quantum computing with closed timelike curves.

DOI
10.48550/arxiv.1101.5355

Related papers

A Qubit, a Coin, and an Advice String Walk Into a Relational Problem

Scott Aaronson, Harry Buhrman, William Kretschmer

Relational problems (those with many possible valid outputs) are different from decision problems, but it is easy to forget just how different. This paper initiates the study of FBQP/qpoly, the class of relational problems solvable in quantum polynomial-time with the help of polynomial-sized quantum advice, along with its analogues for deterministic and randomized computation (FP, FBPP) and advice (/poly, /rpoly). Our first result is that FBQP/qpoly ≠ FBQP/poly, unconditionally, with no oracle - a striking contrast with what we know about the analogous decision classes. The proof repurposes the separation between quantum and classical one-way communication complexities due to Bar-Yossef, Jayram, and Kerenidis. We discuss how this separation raises the prospect of near-term experiments to demonstrate "quantum information supremacy," a form of quantum supremacy that would not depend on unproved complexity assumptions. Our second result is that FBPP ̸ ⊂ FP/poly - that is, Adleman’s Theorem fails for relational problems - unless PSPACE ⊂ NP/poly. Our proof uses IP = PSPACE and time-bounded Kolmogorov complexity. On the other hand, we show that proving FBPP ̸ ⊂ FP/poly will be hard, as it implies a superpolynomial circuit lower bound for PromiseBPEXP. We prove the following further results: - Unconditionally, FP ≠ FBPP and FP/poly ≠ FBPP/poly (even when these classes are carefully defined). - FBPP/poly = FBPP/rpoly (and likewise for FBQP). For sampling problems, by contrast, SampBPP/poly ≠ SampBPP/rpoly (and likewise for SampBQP).

arXiv:2302.10332•Leibniz international proceedings in informatics•Feb 20, 2023•19 citations

Magic coins are useful for small-space quantum machines

A. C. Cem Say, Abuzer Yakaryılmaz

Although polynomial-time probabilistic Turing machines can utilize uncomputable transition probabilities to recognize uncountably many languages with bounded error when allowed to use double logarithmic space, it is known that such “magic coins” give no additional computational power to constant-space versions of those machines. We show that adding a few quantum bits to the model changes the picture dramatically. For every language L, there exists such a two-way quantum finite automaton (2qcfa) that recognizes a language of the same Turing degree as L with bounded error in polynomial time. When used as verifiers in public-coin interactive proof systems, such automata can verify membership in all languages with bounded error, outperforming their classical counterparts, which are known to fail for the palindromes language. Corollaries demonstrate even faster machines when one is allowed to use a counter as memory, and an alternative proof of the uncountability of stochastic languages.

arXiv:1411.7647•Quantum Information and Computation•Sep 1, 2017•13 citations

Classical and quantum computation with small space bounds (PhD thesis)

Abuzer Yakaryılmaz

In this thesis, we introduce a new quantum Turing machine (QTM) model that supports general quantum operators, together with its pushdown, counter, and finite automaton variants, and examine the computational power of classical and quantum machines using small space bounds in many different cases. The main contributions are summarized below. Firstly, we consider QTMs in the unbounded error setting: (i) in some cases of sublogarithmic space bounds, the class of languages recognized by QTMs is shown to be strictly larger than that of classical ones; (ii) in constant space bounds, the same result can still be obtained for restricted QTMs; (iii) the complete characterization of the class of languages recognized by realtime constant space nondeterministic QTMs is given. Secondly, we consider constant space-bounded QTMs in the bounded error setting: (i) we introduce a new type of quantum and probabilistic finite automata (QFAs and PFAs, respectively,) with a special two-way input head which is not allowed to be stationary or move to the left but has the capability to reset itself to its starting position; (ii) the computational power of this type of quantum machine is shown to be superior to that of the probabilistic machine; (iii) based on these models, two-way PFAs and two-way classical-head QFAs are shown to be more succinct than two-way nondeterministic finite automata and their one-way variants; (iv) we also introduce PFAs and QFAs with postselection with their bounded error language classes, and give many characterizations of them. Thirdly, the computational power of realtime QFAs augmented with a write-only memory is investigated by showing many simulation results for different kinds of counter automata. Finally, some lower bounds of realtime classical Turing machines in order to recognize a nonregular language are shown to be tight.

arXiv:1102.0378•Feb 2, 2011•8 citations

No Size-Preserving Amplification with Quantum Advice

Shih-Han Hung, Han-Hsuan Lin

Marriott and Watrous showed that quantum Merlin--Arthur games admit generic error reduction without increasing witness size [Computational Complexity, 2005]. In this work, we show that this state-size-preserving amplification property does not hold for polynomial-time quantum computation with quantum advice. In particular, we present decision problems for which even a vanishing additive error reduction requires longer advice. More precisely, for every polynomially bounded advice length $m(n)=ω(1)$ and every error bound $\varepsilon(n)$ that stays below $1/2$ by at least an inverse polynomial, there is a positive function $δ$ with $δ(n)=O\bigl(\min\{(\log m/m)^{1/4},\ \sqrt{\log m/m}\,/(1/2-\varepsilon(n))\}\bigr)$ such that $\mathsf{BQP}_{\varepsilon}/\mathsf{q}m \subsetneq \mathsf{BQP}_{\varepsilon + δ}/\mathsf{q}m$; for constant $\varepsilon$ the gap is $O(\sqrt{\log m/m})$. Here, $\mathsf{BQP}_\varepsilon/\mathsf{q}m$ is the class of languages recognizable with error at most $\varepsilon(n)$ by a polynomial-time quantum algorithm with an $m(n)$-qubit advice state that only depends on the input length $n$. We show this by proving a stronger separation $\mathsf{P}_{\varepsilon+δ}/\mathsf{r} m \not\subset \mathsf{BQP}_{\varepsilon}/\mathsf{q} m$, where $\mathsf{P}_{\varepsilon}/\mathsf{r}m$ is the class of languages recognizable with error at most $\varepsilon(n)$ by a deterministic polynomial-time algorithm with an $m(n)$-bit advice string sampled from a distribution that depends only on $n$.

quant-pharXiv:2610.03377•Oct 5, 2026•0 citations

The complexity of parameters for probabilistic and quantum computation

Josef Michael Breutzmann

In this dissertation we study some effects of allowing computational models that use parameters whose own computational complexity has a strong effect on the computational complexity of the languages computable from the model. We show that in the probabilistic and quantum models there are parameter sets that allow one to obtain noncomputable outcomes;In Chapter 3 we define BP[beta]P the BPP class based on a coin with bias [beta]. We then show that if [beta] is BPP-computable then it is the case that BP[beta]P = BPP. We also show that each language L in P/CLog is in BP[beta]P for some [beta]. Hence there are some [beta] from which we can compute noncomputable languages. We also examine the robustness of the class BPP with respect to small variations from fairness in the coin;In Chapter 4 we consider measures that are based on polynomial-time computable sequences of biased coins in which the biases are bounded away from both zero and one (strongly positive P-sequences). We show that such a sequence [vector][beta] generates a measure [mu][vector][beta] equivalent to the uniform measure in the sense that if C is a class of languages closed under positive, polynomial-time, truth-table reductions with queries of linear length then C has [mu][vector][beta]-measure zero if and only if it has measure zero relative to the uniform measure [mu]. The classes P, NP, BPP, P/Poly, PH, and PSPACE are among those to which this result applies. Thus the measures of these much-studied classes are robust with respect to changes of this type in the underlying probability measure;In Chapter 5 we introduce the quantum computation model and the quantum complexity class BQP. We claim that the computational complexity of the amplitudes is a critical factor in determining the languages computable using the quantum model. Using results from chapter 3 we show that the quantum model can also compute noncomputable languages from some amplitude sets. Finally, we determine a restriction on the amplitude set to limit the model to the range of languages implicit in others' typical meaning of the class BQP.

•Jan 1, 1996•0 citations

Limitations of quantum advice and one-way communication

Scott Aaronson

Although a quantum state requires exponentially many classical bits to describe, the laws of quantum mechanics impose severe restrictions on how that state can be accessed. This paper shows in three settings that quantum messages have only limited advantages over classical ones. First, we show that BQP/qpoly /spl sube/ PP/poly, where BQP/qpoly is the class of problems solvable in quantum polynomial time, given a polynomial-size "quantum advice state " that depends only on the input length. This resolves a question of Buhrman, and means that we should not hope for an unrelativized separation between quantum and classical advice. Underlying our complexity result is a general relation between deterministic and quantum one-way communication complexities, which applies to partial as well as total functions. Second, we construct an oracle relative to which NP /spl nsub/ BQP/qpoly. To do so, we use the polynomial method to give the first correct proof of a direct product theorem for quantum search. This theorem has many other applications; for example, it can be used to fix and even improve on a flawed result of Klauck about quantum time-space tradeoffs for sorting. Third, we introduce a trace distance method for proving lower bounds on quantum one-way communication complexity. Using this method, we obtain optimal quantum lower bounds for two problems of Ambainis, for which no nontrivial lower bounds were previously known even for classical randomized protocols.

IEEE Conference on Computational Complexity (CCC)•Nov 12, 2004•34 citations
We study the power of classical and quantum algorithms equipped with nonuniform advice, in the form of a coin whose bias encodes useful information. This question takes on particular importance in the quantum case, due to a surprising result that we prove: a quantum finite automaton with just two states can be sensitive to arbitrarily small changes in a coin's bias. This contrasts with classical probabilistic finite automata, whose sensitivity to changes in a coin's bias is bounded by a classic 1970 result of Hellman and Cover. Despite this finding, we are able to bound the power of advice coins for space-bounded classical and quantum computation. We define the classes BPPSPACE/coin and BQPSPACE/coin, of languages decidable by classical and quantum polynomial-space machines with advice coins. Our main theorem is that both classes coincide with PSPACE/poly. Proving this result turns out to require substantial machinery. We use an algorithm due to Neff for finding roots of polynomials in NC; a result from algebraic geometry that lower-bounds the separation of a polynomial's roots; and a result on fixed-points of superoperators due to Aaronson and Watrous, originally proved in the context of quantum computing with closed timelike curves.
Relational problems (those with many possible valid outputs) are different from decision problems, but it is easy to forget just how different. This paper initiates the study of FBQP/qpoly, the class of relational problems solvable in quantum polynomial-time with the help of polynomial-sized quantum advice, along with its analogues for deterministic and randomized computation (FP, FBPP) and advice (/poly, /rpoly). Our first result is that FBQP/qpoly ≠ FBQP/poly, unconditionally, with no oracle - a striking contrast with what we know about the analogous decision classes. The proof repurposes the separation between quantum and classical one-way communication complexities due to Bar-Yossef, Jayram, and Kerenidis. We discuss how this separation raises the prospect of near-term experiments to demonstrate "quantum information supremacy," a form of quantum supremacy that would not depend on unproved complexity assumptions. Our second result is that FBPP ̸ ⊂ FP/poly - that is, Adleman’s Theorem fails for relational problems - unless PSPACE ⊂ NP/poly. Our proof uses IP = PSPACE and time-bounded Kolmogorov complexity. On the other hand, we show that proving FBPP ̸ ⊂ FP/poly will be hard, as it implies a superpolynomial circuit lower bound for PromiseBPEXP. We prove the following further results: - Unconditionally, FP ≠ FBPP and FP/poly ≠ FBPP/poly (even when these classes are carefully defined). - FBPP/poly = FBPP/rpoly (and likewise for FBQP). For sampling problems, by contrast, SampBPP/poly ≠ SampBPP/rpoly (and likewise for SampBQP).
Although polynomial-time probabilistic Turing machines can utilize uncomputable transition probabilities to recognize uncountably many languages with bounded error when allowed to use double logarithmic space, it is known that such “magic coins” give no additional computational power to constant-space versions of those machines. We show that adding a few quantum bits to the model changes the picture dramatically. For every language L, there exists such a two-way quantum finite automaton (2qcfa) that recognizes a language of the same Turing degree as L with bounded error in polynomial time. When used as verifiers in public-coin interactive proof systems, such automata can verify membership in all languages with bounded error, outperforming their classical counterparts, which are known to fail for the palindromes language. Corollaries demonstrate even faster machines when one is allowed to use a counter as memory, and an alternative proof of the uncountability of stochastic languages.
In this thesis, we introduce a new quantum Turing machine (QTM) model that supports general quantum operators, together with its pushdown, counter, and finite automaton variants, and examine the computational power of classical and quantum machines using small space bounds in many different cases. The main contributions are summarized below. Firstly, we consider QTMs in the unbounded error setting: (i) in some cases of sublogarithmic space bounds, the class of languages recognized by QTMs is shown to be strictly larger than that of classical ones; (ii) in constant space bounds, the same result can still be obtained for restricted QTMs; (iii) the complete characterization of the class of languages recognized by realtime constant space nondeterministic QTMs is given. Secondly, we consider constant space-bounded QTMs in the bounded error setting: (i) we introduce a new type of quantum and probabilistic finite automata (QFAs and PFAs, respectively,) with a special two-way input head which is not allowed to be stationary or move to the left but has the capability to reset itself to its starting position; (ii) the computational power of this type of quantum machine is shown to be superior to that of the probabilistic machine; (iii) based on these models, two-way PFAs and two-way classical-head QFAs are shown to be more succinct than two-way nondeterministic finite automata and their one-way variants; (iv) we also introduce PFAs and QFAs with postselection with their bounded error language classes, and give many characterizations of them. Thirdly, the computational power of realtime QFAs augmented with a write-only memory is investigated by showing many simulation results for different kinds of counter automata. Finally, some lower bounds of realtime classical Turing machines in order to recognize a nonregular language are shown to be tight.
Marriott and Watrous showed that quantum Merlin--Arthur games admit generic error reduction without increasing witness size [Computational Complexity, 2005]. In this work, we show that this state-size-preserving amplification property does not hold for polynomial-time quantum computation with quantum advice. In particular, we present decision problems for which even a vanishing additive error reduction requires longer advice. More precisely, for every polynomially bounded advice length m(n)=ω(1)m(n)=ω(1)m(n)=ω(1) and every error bound ε(n)\varepsilon(n)ε(n) that stays below 1/21/21/2 by at least an inverse polynomial, there is a positive function δδδ with δ(n)=O(min⁡{(log⁡m/m)1/4, log⁡m/m /(1/2−ε(n))})δ(n)=O\bigl(\min\{(\log m/m)^{1/4},\ \sqrt{\log m/m}\,/(1/2-\varepsilon(n))\}\bigr)δ(n)=O(min{(logm/m)1/4, logm/m​/(1/2−ε(n))}) such that BQPε/qm⊊BQPε+δ/qm\mathsf{BQP}_{\varepsilon}/\mathsf{q}m \subsetneq \mathsf{BQP}_{\varepsilon + δ}/\mathsf{q}mBQPε​/qm⊊BQPε+δ​/qm; for constant ε\varepsilonε the gap is O(log⁡m/m)O(\sqrt{\log m/m})O(logm/m​). Here, BQPε/qm\mathsf{BQP}_\varepsilon/\mathsf{q}mBQPε​/qm is the class of languages recognizable with error at most ε(n)\varepsilon(n)ε(n) by a polynomial-time quantum algorithm with an m(n)m(n)m(n)-qubit advice state that only depends on the input length nnn. We show this by proving a stronger separation Pε+δ/rm⊄BQPε/qm\mathsf{P}_{\varepsilon+δ}/\mathsf{r} m \not\subset \mathsf{BQP}_{\varepsilon}/\mathsf{q} mPε+δ​/rm⊂BQPε​/qm, where Pε/rm\mathsf{P}_{\varepsilon}/\mathsf{r}mPε​/rm is the class of languages recognizable with error at most ε(n)\varepsilon(n)ε(n) by a deterministic polynomial-time algorithm with an m(n)m(n)m(n)-bit advice string sampled from a distribution that depends only on nnn.
In this dissertation we study some effects of allowing computational models that use parameters whose own computational complexity has a strong effect on the computational complexity of the languages computable from the model. We show that in the probabilistic and quantum models there are parameter sets that allow one to obtain noncomputable outcomes;In Chapter 3 we define BP[beta]P the BPP class based on a coin with bias [beta]. We then show that if [beta] is BPP-computable then it is the case that BP[beta]P = BPP. We also show that each language L in P/CLog is in BP[beta]P for some [beta]. Hence there are some [beta] from which we can compute noncomputable languages. We also examine the robustness of the class BPP with respect to small variations from fairness in the coin;In Chapter 4 we consider measures that are based on polynomial-time computable sequences of biased coins in which the biases are bounded away from both zero and one (strongly positive P-sequences). We show that such a sequence [vector][beta] generates a measure [mu][vector][beta] equivalent to the uniform measure in the sense that if C is a class of languages closed under positive, polynomial-time, truth-table reductions with queries of linear length then C has [mu][vector][beta]-measure zero if and only if it has measure zero relative to the uniform measure [mu]. The classes P, NP, BPP, P/Poly, PH, and PSPACE are among those to which this result applies. Thus the measures of these much-studied classes are robust with respect to changes of this type in the underlying probability measure;In Chapter 5 we introduce the quantum computation model and the quantum complexity class BQP. We claim that the computational complexity of the amplitudes is a critical factor in determining the languages computable using the quantum model. Using results from chapter 3 we show that the quantum model can also compute noncomputable languages from some amplitude sets. Finally, we determine a restriction on the amplitude set to limit the model to the range of languages implicit in others' typical meaning of the class BQP.
Although a quantum state requires exponentially many classical bits to describe, the laws of quantum mechanics impose severe restrictions on how that state can be accessed. This paper shows in three settings that quantum messages have only limited advantages over classical ones. First, we show that BQP/qpoly /spl sube/ PP/poly, where BQP/qpoly is the class of problems solvable in quantum polynomial time, given a polynomial-size "quantum advice state " that depends only on the input length. This resolves a question of Buhrman, and means that we should not hope for an unrelativized separation between quantum and classical advice. Underlying our complexity result is a general relation between deterministic and quantum one-way communication complexities, which applies to partial as well as total functions. Second, we construct an oracle relative to which NP /spl nsub/ BQP/qpoly. To do so, we use the polynomial method to give the first correct proof of a direct product theorem for quantum search. This theorem has many other applications; for example, it can be used to fix and even improve on a flawed result of Klauck about quantum time-space tradeoffs for sorting. Third, we introduce a trace distance method for proving lower bounds on quantum one-way communication complexity. Using this method, we obtain optimal quantum lower bounds for two problems of Ambainis, for which no nontrivial lower bounds were previously known even for classical randomized protocols.