Reinforcement Learning with Quantum Algorithms: Simulating Nematode Behaviour
summary
Uses parameterized quantum circuits as a nematode “brain” trained by policy gradients with a parameter‑shift rule, running on a quantum circuit simulator and on real hardware with error‑suppression. Implements sensory encoding as RX RY RZ rotations, compares against a PyTorch MLP, and reports 12 parameters vs 4,600 for similar performance. Results show quantum runs average ~37 steps with 99% success (classical ~34 steps 100%), but quantum learning is noisier and needs careful initialization.
AI-written summary · about
chapters (16)
- 0:00Quantum nematode simulation intro
- 1:02Simulating C. elegans with quantum
- 2:27Quantum computing primer for RL
- 6:23Parameterized quantum circuits
- 8:29Quantum hardware limitations
- 9:32RL simulation flow explained
- 12:41Classical agent for comparison
- 15:56Quantum agent brain design
- 17:41Running quantum circuits
- 19:49Live quantum agent demo
- 20:55Quantum multiverse worm demo
- 21:40Benchmarking quantum vs classical
- 23:12Key takeaways from results
- 24:57Future directions for QML
- 27:05Quantum ML learning resources
- 28:57Q&A on neural networks
related talks
21mQuantum Reinforcement Learning for Aerospace
19mQuarl: A Learning-Based Quantum Circuit Optimizer
44mQuantum machine learning: Prospects and challenges
27mWhat are quantum computers, and how can we train them in Python?
15mAccelerQ: Accelerating Quantum Eigensolvers With Machine Learning on Quantum Simulators
46mQuantum Machine Learning 101: An Introduction for Beginners with...
31mQuantum neural networks
32mQuantum Artificial Intelligence