1: \begin{abstract}

2: We prove that low-rank matrices can be recovered efficiently from a small number of measurements that are sampled from orbits of a certain matrix group.

3: As a special case, our theory makes statements about the \emph{phase retrieval} problem.

4: Here, the task is to recover a vector given only the \emph{amplitudes} of its inner product with a small number of vectors from an orbit.

5: Variants of the group in question have appeared under different names in many areas of mathematics.

6: In coding theory and quantum information, it is the \emph{complex Clifford group}; in time-frequency analysis the \emph{oscillator group}; and in mathematical physics the \emph{metaplectic group}.

7: It affords one particularly small and highly structured orbit that includes and generalizes the discrete Fourier basis:

8: While the Fourier vectors have coefficients of constant modulus and phases that depend linearly on their index, the vectors in said orbit have phases with a quadratic dependence.

9: In quantum information, the orbit is used extensively and is known as the set of \emph{stabilizer states}.

10: We argue that due to their rich geometric structure and their near-optimal recovery properties, stabilizer states form an ideal model for structured measurements for phase retrieval.

11: Our results hold for $m \geq C  \kappa_r r d \log (d)$ measurements, where the  oversampling factor $\kappa_r$ varies between $\kappa_r=1$ and $\kappa_r=r^2$ depending on the orbit.

12: The reconstruction is stable towards both additive noise and deviations from the assumption of low rank.

13: If the matrices of interest are in addition positive semidefinite, reconstruction may be performed by a simple constrained least squares regression.

14: Our proof methods could be adapted to cover orbits of other groups.

15: \end{abstract}

16: