Backdoor attacks are emerging security threats to deep neural networks.In these attacks,adversaries manipulate the network by constructing training samples embedded with backdoor triggers.The backdoored model performs...Backdoor attacks are emerging security threats to deep neural networks.In these attacks,adversaries manipulate the network by constructing training samples embedded with backdoor triggers.The backdoored model performs as expected on clean test samples but consistently misclassifies samples containing the backdoor trigger as a specific target label.While quantum neural networks(QNNs)have shown promise in surpassing their classical counterparts in certain machine learning tasks,they are also susceptible to backdoor attacks.However,current attacks on QNNs are constrained by the adversary's understanding of the model structure and specific encoding methods.Given the diversity of encoding methods and model structures in QNNs,the effectiveness of such backdoor attacks remains uncertain.In this paper,we propose an algorithm that leverages dataset-based optimization to initiate backdoor attacks.A malicious adversary can embed backdoor triggers into a QNN model by poisoning only a small portion of the data.The victim QNN maintains high accuracy on clean test samples without the trigger but outputs the target label set by the adversary when predicting samples with the trigger.Furthermore,our proposed attack cannot be easily resisted by existing backdoor detection methods.展开更多
Without the geometry of light and logic of photon,observer-observability forms a paradox in modern science,truthequilibrium finds no unification,and mind-light-matter unity is unreachable in spacetime.Subsequently,qua...Without the geometry of light and logic of photon,observer-observability forms a paradox in modern science,truthequilibrium finds no unification,and mind-light-matter unity is unreachable in spacetime.Subsequently,quantum mechanics has been shrouded with mysteries preventing itself from reaching definable causality for a general purpose analytical quantum computing paradigm.Ground-0 Axioms are introduced as an equilibrium-based,dynamic,bipolar set-theoretic unification of the first principles of science and the second law of thermodynamics.Related literatures are critically reviewed to justify the self-evident nature of Ground-0 Axioms.A historical misinterpretation by the founding fathers of quantum mechanics is identified and corrected.That disproves spacetime geometries(including but not limited to Euclidean and Hilbert spaces)as the geometries of light and truth-based logics(including but not limited to bra-ket quantum logic)as the logics of photon.Backed with logically definable causality and Dirac 3-polarizer experiment,bipolar quantum geometry(BQG)and bipolar dynamic logic(BDL)are identified as the geometry of light and the logic of photon,respectively,and wave-particle complementarity is shown less fundamental than bipolar complementarity.As a result,Ground-0 Axioms lead to a geometrical and logical illumination of the quantum and classical worlds as well as the physical and mental worlds.With logical resolutions to the EPR and Schr?dinger’s cat paradoxes,an analytical quantum computing paradigm named quantum intelligence(QI)is introduced.It is shown that QI makes mind-light-matter unity and quantum-digital compatibility logically reachable for quantumneuro-fuzzy AI-machinery with groundbreaking applications.It is contended that Ground-0 Axioms open a new era of science and philosophy—the era of mind-light-matter unity in which humanlevel white-box AI&QI is logically prompted to join Einstein’s grand unification to foster major scientific advances.展开更多
基金supported by the National Natural Science Foundation of China(Grant No.62076042)the National Key Research and Development Plan of China,Key Project of Cyberspace Security Governance(Grant No.2022YFB3103103)the Key Research and Development Project of Sichuan Province(Grant Nos.2022YFS0571,2021YFSY0012,2021YFG0332,and 2020YFG0307)。
文摘Backdoor attacks are emerging security threats to deep neural networks.In these attacks,adversaries manipulate the network by constructing training samples embedded with backdoor triggers.The backdoored model performs as expected on clean test samples but consistently misclassifies samples containing the backdoor trigger as a specific target label.While quantum neural networks(QNNs)have shown promise in surpassing their classical counterparts in certain machine learning tasks,they are also susceptible to backdoor attacks.However,current attacks on QNNs are constrained by the adversary's understanding of the model structure and specific encoding methods.Given the diversity of encoding methods and model structures in QNNs,the effectiveness of such backdoor attacks remains uncertain.In this paper,we propose an algorithm that leverages dataset-based optimization to initiate backdoor attacks.A malicious adversary can embed backdoor triggers into a QNN model by poisoning only a small portion of the data.The victim QNN maintains high accuracy on clean test samples without the trigger but outputs the target label set by the adversary when predicting samples with the trigger.Furthermore,our proposed attack cannot be easily resisted by existing backdoor detection methods.
文摘Without the geometry of light and logic of photon,observer-observability forms a paradox in modern science,truthequilibrium finds no unification,and mind-light-matter unity is unreachable in spacetime.Subsequently,quantum mechanics has been shrouded with mysteries preventing itself from reaching definable causality for a general purpose analytical quantum computing paradigm.Ground-0 Axioms are introduced as an equilibrium-based,dynamic,bipolar set-theoretic unification of the first principles of science and the second law of thermodynamics.Related literatures are critically reviewed to justify the self-evident nature of Ground-0 Axioms.A historical misinterpretation by the founding fathers of quantum mechanics is identified and corrected.That disproves spacetime geometries(including but not limited to Euclidean and Hilbert spaces)as the geometries of light and truth-based logics(including but not limited to bra-ket quantum logic)as the logics of photon.Backed with logically definable causality and Dirac 3-polarizer experiment,bipolar quantum geometry(BQG)and bipolar dynamic logic(BDL)are identified as the geometry of light and the logic of photon,respectively,and wave-particle complementarity is shown less fundamental than bipolar complementarity.As a result,Ground-0 Axioms lead to a geometrical and logical illumination of the quantum and classical worlds as well as the physical and mental worlds.With logical resolutions to the EPR and Schr?dinger’s cat paradoxes,an analytical quantum computing paradigm named quantum intelligence(QI)is introduced.It is shown that QI makes mind-light-matter unity and quantum-digital compatibility logically reachable for quantumneuro-fuzzy AI-machinery with groundbreaking applications.It is contended that Ground-0 Axioms open a new era of science and philosophy—the era of mind-light-matter unity in which humanlevel white-box AI&QI is logically prompted to join Einstein’s grand unification to foster major scientific advances.