arXiv subjects

Changhao Li

Publications and source records attributed to Changhao Li.

At least 37 records · Page 2Linked to original sources

arXiv2026

High Uniformity GaN Micro-pyramids and Platelets by Selective Area Growth

The development of uniform GaN micro-pyramids and platelets via selective area growth is a critical step toward advancing III-nitride device technologies, particularly for micro-light-emitting diode applications. This work investigates the origins of morphological non-uniformity in micro-pyramids and micro-platelets grown by metal-organic chemical vapor deposition (MOCVD). We observe that a direct one-step growth approach leads to significant growth rate inhomogeneity across arrays. To shed light on this issue, we examine the mechanisms driving non-uniformity and explore process modifications aimed at mitigating these effects. Building on these insights, we propose a controlled multi-step growth strategy that combines sequen-tial growth and thermal treatment phases. This approach is demonstrated to enhance surface morphology and structural regularity. The work contributes to the broader objective of enabling scalable, high-precision GaN microstructure fabrication for next-generation optoelectronic applications.

physics.optics↗

arXiv2026

Metriq: A Collaborative Platform for Benchmarking Quantum Computers

The fragmented landscape of quantum computer benchmarks, characterized by system-specific tools and inconsistent evaluation methodologies, hinders reliable cross-platform performance assessment. We introduce Metriq, an open-source collaborative platform for reproducible cross-platform quantum benchmarking that integrates benchmark definition and execution, data collection, and public presentation into a unified workflow. The Metriq benchmark suite spans both system-level metrics that characterize fundamental device properties such as entanglement quality, gate performance, and circuit speed, as well as application-inspired protocols that assess performance on quantum machine learning, optimization, and quantum simulation tasks. Benchmarks are chosen to scale with processor size, and the framework incorporates cost and resource estimation to support practical evaluation. Using Metriq, we collect and publicly release results from more than ten quantum computers across multiple hardware vendors, enabling systematic cross-platform comparison. The resulting curated dataset also reveals the practical strengths and limitations of individual benchmarks, creating a feedback loop that informs the ongoing refinement of the suite. To summarize performance across the benchmark suite, we introduce the Metriq Score, a composite index aggregating benchmark outcomes. We further present cross-benchmark analyses enabled by the shared dataset and their correlations with hardware calibration metrics. Through open development and data sharing, Metriq provides a practical foundation for reproducible benchmarking of quantum computers as hardware and benchmarking methods continue to evolve.

quant-ph↗

arXiv2026

DREAM: Deep Research Evaluation with Agentic Metrics

Deep Research Agents generate analyst-grade reports, yet evaluating them remains challenging due to the absence of a single ground truth and the multidimensional nature of research quality. Recent benchmarks propose distinct methodologies, yet they suffer from the Mirage of Synthesis, where strong surface-level fluency and citation alignment can obscure underlying factual and reasoning defects. We characterize this gap by introducing a taxonomy across four verticals that exposes a critical capability mismatch: static evaluators inherently lack the tool-use capabilities required to assess temporal validity and factual correctness. To address this, we propose DREAM (Deep Research Evaluation with Agentic Metrics), a framework that instantiates the principle of capability parity by making evaluation itself agentic. DREAM structures assessment through an evaluation protocol combining query-agnostic metrics with adaptive metrics generated by a tool-calling agent, enabling temporally aware coverage, grounded verification, and systematic reasoning probes. Controlled evaluations demonstrate DREAM is significantly more sensitive to factual and temporal decay than existing benchmarks, offering a scalable, reference-free evaluation paradigm.

cs.AI↗

arXiv2026

DOS: Dual-Flow Orthogonal Semantic IDs for Recommendation in Meituan

Semantic IDs serve as a key component in generative recommendation systems. They not only incorporate open-world knowledge from large language models (LLMs) but also compress the semantic space to reduce generation difficulty. However, existing methods suffer from two major limitations: (1) the lack of contextual awareness in generation tasks leads to a gap between the Semantic ID codebook space and the generation space, resulting in suboptimal recommendations; and (2) suboptimal quantization methods exacerbate semantic loss in LLMs. To address these issues, we propose Dual-Flow Orthogonal Semantic IDs (DOS) method. Specifically, DOS employs a user-item dual flow-framework that leverages collaborative signals to align the Semantic ID codebook space with the generation space. Furthermore, we introduce an orthogonal residual quantization scheme that rotates the semantic space to an appropriate orientation, thereby maximizing semantic preservation. Extensive offline experiments and online A/B testing demonstrate the effectiveness of DOS. The proposed method has been successfully deployed in Meituan's mobile application, serving hundreds of millions of users.

cs.IR↗

arXiv2026

Entire Chain Uplift Modeling with Context-Enhanced Learning for Intelligent Marketing

Uplift modeling, vital in online marketing, seeks to accurately measure the impact of various strategies, such as coupons or discounts, on different users by predicting the Individual Treatment Effect (ITE). In an e-commerce setting, user behavior follows a defined sequential chain, including impression, click, and conversion. Marketing strategies exert varied uplift effects at each stage within this chain, impacting metrics like click-through and conversion rate. Despite its utility, existing research has neglected to consider the inter-task across all stages impacts within a specific treatment and has insufficiently utilized the treatment information, potentially introducing substantial bias into subsequent marketing decisions. We identify these two issues as the chain-bias problem and the treatment-unadaptive problem. This paper introduces the Entire Chain UPlift method with context-enhanced learning (ECUP), devised to tackle these issues. ECUP consists of two primary components: 1) the Entire Chain-Enhanced Network, which utilizes user behavior patterns to estimate ITE throughout the entire chain space, models the various impacts of treatments on each task, and integrates task prior information to enhance context awareness across all stages, capturing the impact of treatment on different tasks, and 2) the Treatment-Enhanced Network, which facilitates fine-grained treatment modeling through bit-level feature interactions, thereby enabling adaptive feature adjustment. Extensive experiments on public and industrial datasets validate ECUPs effectiveness. Moreover, ECUP has been deployed on the Meituan food delivery platform, serving millions of daily active users, with the related dataset released for future research.

cs.IR↗

arXiv2026

Q-CHOP: Quantum constrained Hamiltonian optimization

Combinatorial optimization problems that arise in science and industry typically have constraints. Yet the presence of constraints makes them challenging to tackle using both classical and quantum optimization algorithms. We propose a new quantum algorithm for constrained optimization, which we call quantum constrained Hamiltonian optimization (Q-CHOP). Our algorithm leverages the observation that for many problems, while the best solution is difficult to find, the worst feasible (constraint-satisfying) solution is known. The basic idea of Q-CHOP is to enforce a Hamiltonian constraint at all times, thereby restricting evolution to the subspace of feasible states, and slowly ``rotate'' an objective Hamiltonian to trace an adiabatic path from the worst feasible state to the best feasible state. Q-CHOP thereby assigns qualitatively distinct roles to the constraint and objective functions of a constrained optimization problem. We additionally propose a version of Q-CHOP that can start in any feasible state. Finally, we benchmark Q-CHOP against the commonly-used adiabatic algorithm of quantum annealing with an objective function that penalizes constraint violation, and find that Q-CHOP consistently performs significantly better on a wide range of problems, including textbook graph problems, knapsack problems, combinatorial auctions, and a real-world financial use case of bond exchange-traded fund basket optimization.

quant-ph↗

arXiv2025

PromptMoE: Generalizable Zero-Shot Anomaly Detection via Visually-Guided Prompt Mixtures

Zero-Shot Anomaly Detection (ZSAD) aims to identify and localize anomalous regions in images of unseen object classes. While recent methods based on vision-language models like CLIP show promise, their performance is constrained by existing prompt engineering strategies. Current approaches, whether relying on single fixed, learnable, or dense dynamic prompts, suffer from a representational bottleneck and are prone to overfitting on auxiliary data, failing to generalize to the complexity and diversity of unseen anomalies. To overcome these limitations, we propose $\mathtt{PromptMoE}$. Our core insight is that robust ZSAD requires a compositional approach to prompt learning. Instead of learning monolithic prompts, $\mathtt{PromptMoE}$ learns a pool of expert prompts, which serve as a basis set of composable semantic primitives, and a visually-guided Mixture-of-Experts (MoE) mechanism to dynamically combine them for each instance. Our framework materializes this concept through a Visually-Guided Mixture of Prompt (VGMoP) that employs an image-gated sparse MoE to aggregate diverse normal and abnormal expert state prompts, generating semantically rich textual representations with strong generalization. Extensive experiments across 15 datasets in industrial and medical domains demonstrate the effectiveness and state-of-the-art performance of $\mathtt{PromptMoE}$.

cs.CV↗

arXiv2025

Matryoshka Pilot: Learning to Drive Black-Box LLMs with LLMs

Despite the impressive generative abilities of black-box large language models (LLMs), their inherent opacity hinders further advancements in capabilities such as reasoning, planning, and personalization. Existing works aim to enhance LLM capabilities via domain-specific adaptation, which require additional training on accessible model parameters, an infeasible option for black-box LLMs. To address this challenge, we introduce Matryoshka Pilot (M-Pilot), a lightweight white-box LLM controller that guides a large-scale black-box LLM generator by decomposing complex tasks into a series of intermediate outputs. Specifically, we consider the black-box LLM as an environment, with M-Pilot serving as a policy to provide intermediate guidance through prompts for driving the black-box LLM. M-Pilot is trained to pivot the outputs of the black-box LLM aligning with preferences during iterative interaction, which enables controllable multi-turn generation and self-improvement in optimizing intermediate guidance. Empirical evaluations on diverse tasks demonstrate that our method effectively enhances the capabilities of black-box LLMs in complex, long-horizon tasks. Our code is publicly available at: https://github.com/lichangh20/Matryoshka.

cs.LG↗

arXiv2025

You Only Evaluate Once: A Tree-based Rerank Method at Meituan

Reranking plays a crucial role in modern recommender systems by capturing the mutual influences within the list. Due to the inherent challenges of combinatorial search spaces, most methods adopt a two-stage search paradigm: a simple General Search Unit (GSU) efficiently reduces the candidate space, and an Exact Search Unit (ESU) effectively selects the optimal sequence. These methods essentially involve making trade-offs between effectiveness and efficiency, while suffering from a severe \textbf{inconsistency problem}, that is, the GSU often misses high-value lists from ESU. To address this problem, we propose YOLOR, a one-stage reranking method that removes the GSU while retaining only the ESU. Specifically, YOLOR includes: (1) a Tree-based Context Extraction Module (TCEM) that hierarchically aggregates multi-scale contextual features to achieve "list-level effectiveness", and (2) a Context Cache Module (CCM) that enables efficient feature reuse across candidate permutations to achieve "permutation-level efficiency". Extensive experiments across public and industry datasets validate YOLOR's performance, and we have successfully deployed YOLOR on the Meituan food delivery platform.

cs.IR↗

arXiv2025

Task-Specific Zero-shot Quantization-Aware Training for Object Detection

Quantization is a key technique to reduce network size and computational complexity by representing the network parameters with a lower precision. Traditional quantization methods rely on access to original training data, which is often restricted due to privacy concerns or security challenges. Zero-shot Quantization (ZSQ) addresses this by using synthetic data generated from pre-trained models, eliminating the need for real training data. Recently, ZSQ has been extended to object detection. However, existing methods use unlabeled task-agnostic synthetic images that lack the specific information required for object detection, leading to suboptimal performance. In this paper, we propose a novel task-specific ZSQ framework for object detection networks, which consists of two main stages. First, we introduce a bounding box and category sampling strategy to synthesize a task-specific calibration set from the pre-trained network, reconstructing object locations, sizes, and category distributions without any prior knowledge. Second, we integrate task-specific training into the knowledge distillation process to restore the performance of quantized detection networks. Extensive experiments conducted on the MS-COCO and Pascal VOC datasets demonstrate the efficiency and state-of-the-art performance of our method. Our code is publicly available at: https://github.com/DFQ-Dojo/dfq-toolkit .

cs.CV↗

arXiv2025

MLE-Dojo: Interactive Environments for Empowering LLM Agents in Machine Learning Engineering

We introduce MLE-Dojo, a Gym-style framework for systematically reinforcement learning, evaluating, and improving autonomous large language model (LLM) agents in iterative machine learning engineering (MLE) workflows. Unlike existing benchmarks that primarily rely on static datasets or single-attempt evaluations, MLE-Dojo provides an interactive environment enabling agents to iteratively experiment, debug, and refine solutions through structured feedback loops. Built upon 200+ real-world Kaggle challenges, MLE-Dojo covers diverse, open-ended MLE tasks carefully curated to reflect realistic engineering scenarios such as data processing, architecture search, hyperparameter tuning, and code debugging. Its fully executable environment supports comprehensive agent training via both supervised fine-tuning and reinforcement learning, facilitating iterative experimentation, realistic data sampling, and real-time outcome verification. Extensive evaluations of eight frontier LLMs reveal that while current models achieve meaningful iterative improvements, they still exhibit significant limitations in autonomously generating long-horizon solutions and efficiently resolving complex errors. Furthermore, MLE-Dojo's flexible and extensible architecture seamlessly integrates diverse data sources, tools, and evaluation protocols, uniquely enabling model-based agent tuning and promoting interoperability, scalability, and reproducibility. We open-source our framework and benchmarks to foster community-driven innovation towards next-generation MLE agents.

cs.LG↗

arXiv2025

MASH: Masked Anchored SpHerical Distances for 3D Shape Representation and Generation

We introduce Masked Anchored SpHerical Distances (MASH), a novel multi-view and parametrized representation of 3D shapes. Inspired by multi-view geometry and motivated by the importance of perceptual shape understanding for learning 3D shapes, MASH represents a 3D shape as a collection of observable local surface patches, each defined by a spherical distance function emanating from an anchor point. We further leverage the compactness of spherical harmonics to encode the MASH functions, combined with a generalized view cone with a parameterized base that masks the spatial extent of the spherical function to attain locality. We develop a differentiable optimization algorithm capable of converting any point cloud into a MASH representation accurately approximating ground-truth surfaces with arbitrary geometry and topology. Extensive experiments demonstrate that MASH is versatile for multiple applications including surface reconstruction, shape generation, completion, and blending, achieving superior performance thanks to its unique representation encompassing both implicit and explicit features.

cs.CV↗

arXiv2025

Des-q: a quantum algorithm to provably speedup retraining of decision trees

Decision trees are widely adopted machine learning models due to their simplicity and explainability. However, as training data size grows, standard methods become increasingly slow, scaling polynomially with the number of training examples. In this work, we introduce Des-q, a novel quantum algorithm to construct and retrain decision trees for regression and binary classification tasks. Assuming the data stream produces small, periodic increments of new training examples, Des-q significantly reduces the tree retraining time. Des-q achieves a logarithmic complexity in the combined total number of old and new examples, even accounting for the time needed to load the new samples into quantum-accessible memory. Our approach to grow the tree from any given node involves performing piecewise linear splits to generate multiple hyperplanes, thus partitioning the input feature space into distinct regions. To determine the suitable anchor points for these splits, we develop an efficient quantum-supervised clustering method, building upon the q-means algorithm introduced by Kerenidis et al. We benchmark the simulated version of Des-q against the state-of-the-art classical methods on multiple data sets and observe that our algorithm exhibits similar performance to the state-of-the-art decision trees while significantly speeding up the periodic tree retraining.

quant-ph↗

arXiv2025

A Comprehensive Study of Structural Pruning for Vision Models

Structural pruning has emerged as a promising approach for producing more efficient models. Nevertheless, the community suffers from a lack of standardized benchmarks and metrics, leaving the progress in this area not fully comprehended. To fill this gap, we present the first comprehensive benchmark, termed PruningBench, for structural pruning. PruningBench showcases the following three characteristics: 1) PruningBench employs a unified and consistent framework for evaluating the effectiveness of diverse structural pruning techniques; 2) PruningBench systematically evaluates 16 existing pruning methods, encompassing a wide array of models (e.g., CNNs and ViTs) and tasks (e.g., classification and detection); 3) PruningBench provides easily implementable interfaces to facilitate the implementation of future pruning methods, and enables the subsequent researchers to incorporate their work into our leaderboards. We provide an online pruning platform for customizing pruning tasks and reproducing all results in this paper. Leaderboard results can also be available.

cs.AI↗

arXiv2024

Electrostatic self-assembly of neutral particles on a dielectric substrate: A theoretical study via a multiple-image method and an effective-dipole approach

A multiple-image method is developed to accurately calculate the electrostatic interaction between neutral dielectric particles and a uniformly charged dielectric substrate. The difference in dielectric constants between the particle and the solvent medium leads to a reversal of positive and negative polarizations in the particle. The variance in dielectric constants between the solvent medium and the substrate causes a transition from attractive to repulsive forces between the particle and the substrate. A nonuniform electrostatic field is generated by the polarized charges on the substrate due to mutual induction. These characteristics of electrostatic manipulation determine whether particles are adsorbed onto the substrate or pushed away from it. The self-assembled particles tend to aggregate in a stable hexagonal structure on the substrate. These findings provide new insights into self-assembly processes involving neutral particles on a dielectric substrate.

cond-mat.soft↗

arXiv2024

Non-Hermitian Spin-Spin Interaction Mediated by Chiral Phonons

Non-Hermiticity and chirality are two fundamental properties known to give rise to various intriguing phenomena. However, the interplay between these properties has been rarely explored. In this work, we bridge this gap by introducing an off-diagonal non-Hermitian spin-spin interaction mediated by chiral phonons. This interaction arises from the spin-selectivity due to the locking between phonon momentum and angular momentum in chiral materials. The resulting non-Hermitian interaction mediated by the vacuum field of chiral phonons can reach the kHz range for electron spins and can be further enhanced by externally driven mechanical waves, potentially leading to observable effects in the quantum regime. Moreover, the long-range nature of phonon-mediated interactions enables the realization of the long-desired non-Hermitian interaction among multiple spins. The effect proposed in this work may have wide-ranging applications in cascaded quantum systems, non-Hermitian many-body physics, and non-Hermitian cooling.

quant-ph↗

arXiv2024

Parameter Setting Heuristics Make the Quantum Approximate Optimization Algorithm Suitable for the Early Fault-Tolerant Era

Quantum Approximate Optimization Algorithm (QAOA) is one of the most promising quantum heuristics for combinatorial optimization. While QAOA has been shown to perform well on small-scale instances and to provide an asymptotic speedup over state-of-the-art classical algorithms for some problems, fault-tolerance is understood to be required to realize this speedup in practice. The low resource requirements of QAOA make it particularly suitable to benchmark on early fault-tolerant quantum computing (EFTQC) hardware. However, the performance of QAOA depends crucially on the choice of the free parameters in the circuit. The task of setting these parameters is complicated in the EFTQC era by the large overheads, which preclude extensive classical optimization. In this paper, we summarize recent advances in parameter setting in QAOA and show that these advancements make EFTQC experiments with QAOA practically viable.

quant-ph↗

arXiv2024

Efficient Quantum Transduction Using Anti-Ferromagnetic Topological Insulators

Transduction of quantum information between distinct quantum systems is an essential step in various applications, including quantum communications and quantum computing. However, mediating photons of vastly different frequencies and designing high-performance transducers are highly nontrivial, due to multifaceted and sometimes conflicting requirements. In this work, we first discuss some general principles for quantum transducer design, and then propose solid-state anti-ferromagnetic topological insulators to serve as particularly effective transducers. First, the anti-ferromagnetic order can minimize detrimental influences on nearby quantum systems caused by magnetic interactions. Second, topological insulators exhibit band-inversion, which can greatly enhance their optical responses. This property, coupled with robust spin-orbit coupling and high spin density, results in strong nonlinear interaction in magnetic topological insulators, thereby substantially improving transduction efficiency. Using MnBi2Te4 as an example, we discuss the potential experimental realization of quantum transduction based on magnetic topological materials. Particularly, we showcase that quantum transduction efficiency exceeding 90% can be achieved with modest experimental requirements, while the transduction bandwidth can reach the GHz range. The strong nonlinear photonic interactions in magnetic topological insulators can find diverse applications besides quantum transduction, such as quantum squeezing.

cond-mat.mtrl-sci↗