Quantum adaptive agents with efficient long-term memories
arXiv:2108.10876 · doi:10.1103/PhysRevX.12.011007
Abstract
Central to the success of adaptive systems is their ability to interpret signals from their environment and respond accordingly -- they act as agents interacting with their surroundings. Such agents typically perform better when able to execute increasingly complex strategies. This comes with a cost: the more information the agent must recall from its past experiences, the more memory it will need. Here we investigate the power of agents capable of quantum information processing. We uncover the most general form a quantum agent need adopt to maximise memory compression advantages, and provide a systematic means of encoding their memory states. We show these encodings can exhibit extremely favourable scaling advantages relative to memory-minimal classical agents, particularly when information must be retained about events increasingly far into the past.
arXiv comment16 pages, 4 figures
References in corpus (40)
- Quantum autoencoders for efficient compression of quantum data
- Computational Mechanics: Pattern and Prediction, Structure and Simplicity
- Quantum-enhanced machine learning
- Quantum reinforcement learning
- Quantum channels and memory effects
- Quantum generalisation of feedforward neural networks
- Quantum speedup for active learning agents
- Occam's Quantum Razor: How Quantum Mechanics can reduce the complexity of classical models
- Experimental Realization of a Quantum Autoencoder: The Compression of Qutrits via Machine Learning
- Quantum random walks with history dependence
- A practical, unitary simulator for non-Markovian complex processes
- Quantum Data Compression of a Qubit Ensemble
- Realization of a quantum autoencoder for lossless compression of quantum data
- Computational Mechanics of Input-Output Processes: Structured transformations and the -transducer
- Extreme dimensionality reduction with quantum modelling
- Optimal stochastic modelling with unitary quantum dynamics
- Quantum walks with memory - goldfish, elephants and wise old men
- Optimal classical simulation of state-independent quantum contextuality
- Thermodynamical cost of some interpretations of quantum theory
- Informational and Causal Architecture of Discrete-Time Renewal Processes
- Using quantum theory to reduce the complexity of input-output processes
- Informational and Causal Architecture of Continuous-time Renewal and Hidden Semi-Markov Processes
- Extreme Quantum Advantage for Rare-Event Sampling
- Superior memory efficiency of quantum devices for the simulation of continuous-time stochastic processes
- Leveraging Environmental Correlations: The Thermodynamics of Requisite Variety
- The classical-quantum divergence of complexity in modelling spin chains
- Causal Asymmetry in a Quantum World
- Interfering trajectories in experimental quantum-enhanced stochastic simulation
- Optimized Bacteria are Environmental Prediction Engines
- Quantum Stopwatch: How To Store Time in a Quantum Memory
- Memory compression and thermal efficiency of quantum implementations of non-deterministic hidden Markov models
- Strong and Weak Optimizations in Classical and Quantum Models of Stochastic Processes
- Single-shot quantum memory advantage in the simulation of stochastic processes
- Memory-efficient tracking of complex temporal and symbolic dynamics with quantum simulators
- Thermal Efficiency of Quantum Memory Compression
- Robust inference of memory structure for efficient quantum modelling of stochastic processes
- Quantum clocks are more precise than classical ones
- Quantum coarse-graining for extreme dimension reduction in modelling stochastic temporal dynamics
- Ultimate limit on time signal generation
- Measures of distinguishability between stochastic processes