This lecture provides a comprehensive overview of quantum chaos, starting with single-particle systems and progressing to many-body phenomena. The speaker begins by discussing the role of symmetries, particularly time-reversal symmetry, which dictates whether eigenstates are real or complex. This leads to a statistical description of chaotic eigenstates as "random states," characterized by their localization properties via the inverse participation ratio (IPR). The core of the talk introduces Random Matrix Theory (RMT) as the statistical model for the energy levels of chaotic systems, explaining how ensembles like the Gaussian Orthogonal (GOE) and Gaussian Unitary (GUE) predict universal spectral features, most notably level repulsion, which is contrasted with the Poisson statistics of integrable systems. The lecture then extends these concepts to many-body systems, exploring how quantum chaos manifests as the generation of nearly maximal multipartite entanglement, consistent with predictions for random states (the Page value). The principle of entanglement monogamy is used to explain why bipartite entanglement between individual components (e.g., qubits) is suppressed in chaotic regimes. The talk concludes by discussing advanced diagnostics like the spectral form factor and out-of-time-ordered correlators (OTOCs).
Statistical Signatures of Quantum Chaos: From Random States to Many-Body Entanglement
0:01:19 Time-Reversal Symmetry and Eigenstates: The lecture begins by explaining anti-unitary symmetries, like time-reversal. A key takeaway is that systems with time-reversal symmetry have eigenstates that can be represented as purely real vectors in an appropriate basis.
0:06:22 Random States as Models for Chaos: Chaotic eigenstates are modeled as random states, where the amplitudes are uniformly distributed on a sphere subject only to normalization.
0:10:08 Inverse Participation Ratio (IPR): The IPR is introduced as a measure of a state's localization. For chaotic systems with time-reversal symmetry (real states), the participation ratio is N/3, meaning about one-third of the basis states are significant. For systems without this symmetry (complex states), it is N/2, indicating greater delocalization.
0:22:59 Level Repulsion in Chaotic Systems: A hallmark of quantum chaos is level repulsion, where energy levels avoid crossing. This is contrasted with integrable systems, where energy levels are uncorrelated and follow Poisson statistics, leading to level clustering.
0:39:03 Introduction to Random Matrix Theory (RMT): RMT is presented as the statistical framework for quantum chaos. The Gaussian Orthogonal Ensemble (GOE) models time-reversal symmetric systems (real symmetric matrices), while the Gaussian Unitary Ensemble (GUE) models systems without it (complex Hermitian matrices).
0:45:04 Origin of Level Repulsion in RMT: The joint probability distribution of eigenvalues in RMT includes a Vandermonde determinant term (product of eigenvalue differences), which explicitly suppresses the probability of finding two eigenvalues close together, thus causing repulsion.
0:59:18 The Spectral Form Factor (SFF): The SFF is a two-point correlation function of the energy spectrum. For chaotic systems, it exhibits a characteristic "ramp" and "plateau" structure, distinct from the flat SFF of integrable (Poisson) systems. It is noted to be a non-self-averaging quantity, requiring an ensemble average to reveal its universal form.
1:10:53 Quantum Chaos in Many-Body Systems: The concepts are extended to many-body physics, where chaos is linked to thermalization and entanglement.
1:21:23 Chaos Creates High Entanglement: Chaotic many-body systems evolve to produce nearly maximally entangled states. The average entanglement of a random state approaches the "Page value" (log N - 1/2), which is very close to the maximum possible entanglement.
1:27:00 Monogamy of Entanglement: This principle states that if two subsystems are maximally entangled with each other, they cannot be entangled with anything else. In a chaotic many-body system, entanglement becomes highly multipartite (spread across many parts), which suppresses bipartite entanglement between any two individual particles (e.g., two qubits).
1:43:41 Out-of-Time-Ordered Correlators (OTOCs): Discussed in the Q&A, OTOCs are a tool to measure quantum chaos and information scrambling. In systems with a classical limit, the OTOC can show exponential growth, defining a quantum Lyapunov exponent.