arXiv subjects

Hansen Wang

Publications and source records attributed to Hansen Wang.

4 recordsLinked to original sources

arXiv2022

A Realizable GAS-based Quantum Algorithm for Traveling Salesman Problem

The paper proposes a quantum algorithm for the traveling salesman problem (TSP) based on the Grover Adaptive Search (GAS), which can be successfully executed on IBM's Qiskit library. Under the GAS framework, there are at least two fundamental difficulties that limit the application of quantum algorithms for combinatorial optimization problems. One difficulty is that the solutions given by the quantum algorithms may not be feasible. The other difficulty is that the number of qubits of current quantum computers is still very limited, and it cannot meet the minimum requirements for the number of qubits required by the algorithm. In response to the above difficulties, we designed and improved the Hamiltonian Cycle Detection (HCD) oracle based on mathematical theorems. It can automatically eliminate infeasible solutions during the execution of the algorithm. On the other hand, we design an anchor register strategy to save the usage of qubits. The strategy fully considers the reversibility requirement of quantum computing, overcoming the difficulty that the used qubits cannot be simply overwritten or released. As a result, we successfully implemented the numerical solution to TSP on IBM's Qiskit. For the seven-node TSP, we only need 31 qubits, and the success rate in obtaining the optimal solution is 86.71%.

quant-ph↗

arXiv2022

The Homogenization Cone: Polar Cone and Projection

Let $C$ be a closed convex subset of a real Hilbert space containing the origin, and assume that $K$ is the homogenization cone of $C$, i.e., the smallest closed convex cone containing $C \times \{1\}$. Homogenization cones play an important role in optimization as they include, for instance, the second-order/Lorentz/"ice cream" cone. In this note, we discuss the polar cone of $K$ as well as an algorithm for finding the projection onto $K$ provided that the projection onto $C$ is available. Various examples illustrate our results.

math.OC↗

arXiv2021

Learning to Deceive Knowledge Graph Augmented Models via Targeted Perturbation

Knowledge graphs (KGs) have helped neural models improve performance on various knowledge-intensive tasks, like question answering and item recommendation. By using attention over the KG, such KG-augmented models can also "explain" which KG information was most relevant for making a given prediction. In this paper, we question whether these models are really behaving as we expect. We show that, through a reinforcement learning policy (or even simple heuristics), one can produce deceptively perturbed KGs, which maintain the downstream performance of the original KG while significantly deviating from the original KG's semantics and structure. Our findings raise doubts about KG-augmented models' ability to reason about KG information and give sensible explanations.

cs.CL↗

arXiv2019

Cryo-EM structures of atomic surfaces and host-guest chemistry in metal-organic frameworks

Host-guest interactions govern the chemistry of a broad range of functional materials. However, the weak bonding between host and guest prevents atomic-resolution studies of their structure and chemistry using transmission electron microscopy (TEM). This problem is exacerbated in metal-organic frameworks (MOF), in which the host framework is easily damaged by the electron beam. Here, we use cryogenic-electron microscopy (cryo-EM) to simultaneously address these two challenges and resolve the atomic surface structure of zeolitic imidazolate framework (ZIF-8) and its interaction with guest CO2 molecules. We image atomic step-edge sites on the ZIF-8 surface that provides possible insight to its growth behavior. Furthermore, we observe two distinct binding sites for CO2 within the ZIF-8 pore, which are both predicted by density functional theory (DFT) to be energetically favorable. This CO2 insertion induces an apparent ~3% lattice expansion along the <002> and <011> directions of the ZIF-8 unit cell. The ability to stabilize MOFs and preserve their host-guest chemistry opens a rich materials space for scientific exploration and discovery using cryo-EM.

cond-mat.mtrl-sci↗