arXiv Science⌕ Search

arXiv subjects

Michelle Xu

Publications and source records attributed to Michelle Xu.

6 recordsLinked to original sources

Black holes as black boxes: a query lower bound on python's lunch decoding

The python's lunch conjecture relates the complexity of bulk reconstruction in holographic tensor networks to an exponential cost associated with post-selection. We establish a tight query lower bound for this decoding problem in a Haar random oracle model, with a python's lunch circuit that applies Haar random unitary $U$ to a logical input and fixed ancillas, followed by post-selection on $m$ output qubits. Given one copy of the normalized encoded state and coherent oracle access to $U$ and $U^\dagger$, including controlled queries, we prove that recovering even a single logical bit with constant average advantage requires $Ω(2^{m/2})$ queries. Together with the state-independent amplitude amplification upper bound for approximately isometric encodings, this establishes query complexity $Θ(2^{m/2})$ at constant decoding accuracy. We also comment on the decoding cost for other "admissible'" ensembles of unitaries $U$ which give rise to families of approximately isometric holographic maps.

quant-ph↗

Deep thermalization with and without quantum chaos

Can measuring part of a quantum system make the remainder more random than the whole? Deep thermalization is a phenomenon in which measurement on part of a thermalizing quantum many-body system produces a subsystem ensemble of states whose higher moments approximate those of Haar-random states, i.e. a quantum state design. In this work, we tightly limit how much randomness from the global state can be transferred to the subsystem in this way: without further assumptions, a global k-design degrades to at most a k/2-design on the unmeasured subsystem. We show this optimality bound by constructing a global k-design for which each sampled state hides a sign preference in its projected distribution that can always be discovered by analyzing moments of its projected ensemble greater than k/2. Strikingly, we show this design loss can be reversed when the global design is generated by chaotic dynamics. Under the chaotic constant-time evolution of a Gaussian Unitary Ensemble (GUE) Hamiltonian, we find that the projected ensemble becomes exactly Haar-random in the thermodynamic limit --- even if the global state seems to be only a $\mathcal{O}(1)$-design. Moreover, we show this phenomenon does not require the full GUE structure: it persists under substantially weaker assumptions on the spectrum and eigenbasis, and admits finite-size versions with increasingly accurate projected designs as the measured subsystem grows. Our results identify global chaotic dynamics as a mechanism by which measurements can ``concentrate'' quantum randomness and place tight limits on when such amplification is possible.

quant-ph↗

LLM Router: Rethinking Routing with Prefill Activations

Existing routers rely on semantic query features or handcrafted features, which often fail to capture model-specific failures or intrinsic task difficulty. We instead route using internal LLM activations, specifically the residual stream. Our key idea, Encoder-Target Decoupling, separates the model that produces the predictive signal (the Encoder) from the model whose correctness is being estimated (the Target), allowing open-weight encoders to predict the performance of closed-source target models. We evaluate layerwise geometric probes, finding that Fisher Separability ($J$) effectively identifies informative layers, supported by Effective Dimensionality ($d_{\mathrm{eff}}$) diagnostics. We then utilize a SharedTrunkNet, a joint multi-output MLP that predicts simultaneous correctness probabilities across candidate models using concatenated prefill features. In our experiments, SharedTrunkNet consistently outperforms semantic baselines. At its best, SharedTrunkNet closes 45.58% of the gap between the strongest standalone model and the oracle while achieving 74.31% cost savings relative to the most expensive model. These results demonstrate that prefill activations provide a robust routing signal, establishing activation-based routing as a high-performance alternative to purely semantic selection.

cs.CL↗

Cryptographic tests of the python's lunch conjecture

In the AdS/CFT correspondence, a subregion of the CFT allows for the recovery of a corresponding subregion of the bulk known as its entanglement wedge. In some cases, an entanglement wedge contains a locally but not globally minimal surface homologous to the CFT subregion, in which case it is said to contain a python's lunch. It has been proposed that python's lunch geometries should be modelled by tensor networks that feature projective operations where the wedge narrows. This model leads to the python's lunch (PL) conjecture, which asserts that reconstructing information from past the locally minimal surface is computationally difficult. In this work, we use cryptographic tools related to a primitive known as the Conditional Disclosure of Secrets (CDS) to develop consequences of the projective tensor network model that can be checked directly in AdS/CFT. We argue from the tensor network picture that the mutual information between appropriate CFT subregions is lower bounded linearly by an area difference associated with the geometry of the lunch. Recalling that the mutual information is also computed by bulk extremal surfaces, this gives a checkable geometrical consequence of the tensor network model. We prove weakened versions of this geometrical statement in asymptotically AdS$_{2+1}$ spacetimes satisfying the null energy condition, and confirm it in some example geometries, supporting the tensor network model and by proxy the PL conjecture.

hep-th↗

Efficient Unitary T-designs from Random Sums

Unitary $T$-designs play an important role in quantum information, with diverse applications in quantum algorithms, benchmarking, tomography, and communication. Until now, the most efficient construction of unitary $T$-designs for $n$-qudit systems has been via random local quantum circuits, which have been shown to converge to approximate $T$-designs in the diamond norm using $O(T^{5+o(1)} n^2)$ quantum gates. In this work, we provide a new construction of $T$-designs via random matrix theory using $\tilde{O}(T^2 n^2)$ quantum gates. Our construction leverages two key ideas. First, in the spirit of central limit theorems, we approximate the Gaussian Unitary Ensemble (GUE) by an i.i.d. sum of random Hermitian matrices. Second, we show that the product of just two exponentiated GUE matrices is already approximately Haar random. Thus, multiplying two exponentiated sums over rather simple random matrices yields a unitary $T$-design, via Hamiltonian simulation. A central feature of our proof is a new connection between the polynomial method in quantum query complexity and the large-dimension ($N$) expansion in random matrix theory. In particular, we show that the polynomial method provides exponentially improved bounds on the high moments of certain random matrix ensembles, without requiring intricate Weingarten calculations. In doing so, we define and solve a new type of moment problem on the unit circle, asking whether a finite number of equally weighted points, corresponding to eigenvalues of unitary matrices, can reproduce a given set of moments.

quant-ph↗

Non-local computation and the black hole interior

In a two sided black hole, systems falling in from opposite asymptotic regions can meet inside the black hole and interact. This is the case even while the two CFTs describing each asymptotic region are non-interacting. Here, we relate these behind the horizon interactions to non-local quantum computations. This gives a quantum circuit perspective on these interactions, which applies whenever the interaction occurs in the past of a certain extremal surface that sits inside the black hole and in arbitrary dimension. Whenever our perspective applies, we obtain a boundary signature for these interior collisions which is stated in terms of the mutual information. We further revisit the connection discussed earlier between bulk interactions in one sided AdS geometries and non-local computation, and recycle some of our techniques to offer a new perspective on making that connection precise.

hep-th↗