Two hundred years BCE, the ancient Greeks were able to estimate astronomical positions and eclipses with the Antikythera mechanism, a device that mechanically captured aspects of the logic of planetary orbits - an early concept of an analog computer.

Today, we can extend this idea into the quantum realm.

Continuous-variable quantum platforms offer a promising route toward analog quantum simulation. Much like the Antikythera mechanism, a carefully engineered laboratory system can emulate the dynamics of a more complex target system. In particular, Gaussian boson samplers are promising candidates for demonstrating near-term computational quantum advantage. However, experimental imperfections and incomplete knowledge of the underlying quantum states can compromise this advantage, distort the simulated dynamics, and raise questions about whether the resulting evolution remains consistent with a physically admissible model of the target system. Characterizing, modeling, and quantifying these effects is therefore essential for establishing the regimes in which quantum simulations remain physically meaningful and quantum advantage can be reliably demonstrated. Since classical simulation becomes computationally intractable precisely in these regimes, I am also interested in developing efficiently computable witnesses that can assess the physical consistency and reliability of quantum simulators without requiring their full classical simulation.