Searcharxiv⌕ Search

arXiv subjects

Haibin Wang

Publications and source records attributed to Haibin Wang.

At least 19 recordsLinked to original sources
arXiv2026

Eigenvalues on spheres

For every smooth Riemannian metric on the two sphere whose Gaussian curvature is bounded below by one, we prove that each positive Laplace eigenvalue, counted with multiplicity, is no smaller than the corresponding eigenvalue of the unit round sphere. Equality at any positive position in the ordered spectrum forces the metric to be isometric to the unit round metric. We further establish a sharp finite spectral counting comparison for Alexandrov two spheres with curvature bounded below by one. At every positive spectral threshold of the unit round sphere, the number of Laplace eigenvalues below or at that threshold, counted with multiplicity, does not exceed the corresponding number for the round sphere. Equality at any such threshold forces the Alexandrov sphere to be isometric to the unit round sphere. As an application, we obtain the sharp Euclidean dimension bound for spaces of polynomial growth harmonic functions on complete three dimensional manifolds with nonnegative sectional curvature and positive asymptotic volume ratio, together with rigidity in the equality case.

math.DG↗
arXiv2026

Uniform estimates and Brezis-Merle type inequalities for the $k$-Hessian equation

In this paper, we prove a Brezis-Merle type inequality for $k$-convex functions vanishing on the boundary. As an application, we establish an Alexandrov-Bakelman-Pucci type estimate for the intermediate Hessian equation. Furthermore, we establish a concentration-compactness principle for the blow-up behavior of solutions to the mean field type $k$-Hessian equation.

math.AP↗
arXiv2026

LongCat-Flash-Thinking-2601 Technical Report

We introduce LongCat-Flash-Thinking-2601, a 560-billion-parameter open-source Mixture-of-Experts (MoE) reasoning model with superior agentic reasoning capability. LongCat-Flash-Thinking-2601 achieves state-of-the-art performance among open-source models on a wide range of agentic benchmarks, including agentic search, agentic tool use, and tool-integrated reasoning. Beyond benchmark performance, the model demonstrates strong generalization to complex tool interactions and robust behavior under noisy real-world environments. Its advanced capability stems from a unified training framework that combines domain-parallel expert training with subsequent fusion, together with an end-to-end co-design of data construction, environments, algorithms, and infrastructure spanning from pre-training to post-training. In particular, the model's strong generalization capability in complex tool-use are driven by our in-depth exploration of environment scaling and principled task construction. To optimize long-tailed, skewed generation and multi-turn agentic interactions, and to enable stable training across over 10,000 environments spanning more than 20 domains, we systematically extend our asynchronous reinforcement learning framework, DORA, for stable and efficient large-scale multi-environment training. Furthermore, recognizing that real-world tasks are inherently noisy, we conduct a systematic analysis and decomposition of real-world noise patterns, and design targeted training procedures to explicitly incorporate such imperfections into the training process, resulting in improved robustness for real-world applications. To further enhance performance on complex reasoning tasks, we introduce a Heavy Thinking mode that enables effective test-time scaling by jointly expanding reasoning depth and width through intensive parallel thinking.

cs.AI↗
arXiv2024

Data-Juicer 2.0: Cloud-Scale Adaptive Data Processing for and with Foundation Models

Foundation models demand advanced data processing for their vast, multimodal datasets. However, traditional frameworks struggle with the unique complexities of multimodal data. In response, we present Data-Juicer 2.0, a data processing system backed by 100+ data processing operators spanning text, image, video, and audio modalities, supporting more critical tasks including data analysis, synthesis, annotation, and foundation model post-training. With seamless compatibility and dedicated optimization for popular dataset hubs like Hugging Face and computing engines like Ray, it improves upon its predecessor in terms of usability, efficiency, and programmability. It features an easily accessible user interface layer that supports decoupled Python interactions, RESTful APIs, and conversational commands. Its new runtime layer offers adaptive execution across diverse scales and environments, abstracting away system complexities. Extensive empirical evaluations demonstrate Data-Juicer 2.0's remarkable performance and scalability, highlighting its capability to efficiently process TB-level data with 10k+ CPU cores. The system is publicly available and has been widely adopted in diverse research fields and real-world products such as Alibaba Cloud PAI. We actively maintain the system and share practical insights to foster research and applications of next-generation foundation models.

cs.DC↗
arXiv2024

Data-Juicer Sandbox: A Feedback-Driven Suite for Multimodal Data-Model Co-development

The emergence of multimodal large models has advanced artificial intelligence, introducing unprecedented levels of performance and functionality. However, optimizing these models remains challenging due to historically isolated paths of model-centric and data-centric developments, leading to suboptimal outcomes and inefficient resource utilization. In response, we present a new sandbox suite tailored for integrated data-model co-development. This sandbox provides a feedback-driven experimental platform, enabling cost-effective iteration and guided refinement of both data and models. Our proposed ``Probe-Analyze-Refine'' workflow, validated through practical use cases on multimodal tasks such as image-text pre-training with CLIP, image-to-text generation with LLaVA-like models, and text-to-video generation with DiT-based models, yields transferable and notable performance boosts, such as topping the VBench leaderboard. A comprehensive set of over 100 experiments demonstrated the suite's usability and extensibility, while also uncovering insights into the interplay between data quality, diversity, model behavior, and computational costs. All codes, datasets, and models are open-sourced to foster future research and applications that would otherwise be infeasible due to the lack of a dedicated co-development infrastructure.

cs.AI↗
arXiv2024

UniDM: A Unified Framework for Data Manipulation with Large Language Models

Designing effective data manipulation methods is a long standing problem in data lakes. Traditional methods, which rely on rules or machine learning models, require extensive human efforts on training data collection and tuning models. Recent methods apply Large Language Models (LLMs) to resolve multiple data manipulation tasks. They exhibit bright benefits in terms of performance but still require customized designs to fit each specific task. This is very costly and can not catch up with the requirements of big data lake platforms. In this paper, inspired by the cross-task generality of LLMs on NLP tasks, we pave the first step to design an automatic and general solution to tackle with data manipulation tasks. We propose UniDM, a unified framework which establishes a new paradigm to process data manipulation tasks using LLMs. UniDM formalizes a number of data manipulation tasks in a unified form and abstracts three main general steps to solve each task. We develop an automatic context retrieval to allow the LLMs to retrieve data from data lakes, potentially containing evidence and factual information. For each step, we design effective prompts to guide LLMs to produce high quality results. By our comprehensive evaluation on a variety of benchmarks, our UniDM exhibits great generality and state-of-the-art performance on a wide variety of data manipulation tasks.

cs.AI↗
arXiv2023

Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation

Large language models (LLMs) have emerged as a new paradigm for Text-to-SQL task. However, the absence of a systematical benchmark inhibits the development of designing effective, efficient and economic LLM-based Text-to-SQL solutions. To address this challenge, in this paper, we first conduct a systematical and extensive comparison over existing prompt engineering methods, including question representation, example selection and example organization, and with these experimental results, we elaborate their pros and cons. Based on these findings, we propose a new integrated solution, named DAIL-SQL, which refreshes the Spider leaderboard with 86.6% execution accuracy and sets a new bar. To explore the potential of open-source LLM, we investigate them in various scenarios, and further enhance their performance with supervised fine-tuning. Our explorations highlight open-source LLMs' potential in Text-to-SQL, as well as the advantages and disadvantages of the supervised fine-tuning. Additionally, towards an efficient and economic LLM-based Text-to-SQL solution, we emphasize the token efficiency in prompt engineering and compare the prior studies under this metric. We hope that our work provides a deeper understanding of Text-to-SQL with LLMs, and inspires further investigations and broad applications.

cs.DB↗
arXiv2023

An Improved Quantum Private Set Intersection Protocol Based on Hadamard Gates

Recently, Liu and Yin (Int. J. Theor. Phys. 60, 2074-2083 (2021)) proposed a two-party private set intersection protocol based on quantum Fourier transform. We find the participant can deduce the other party's private information, which violates the security requirement of private set computation. In order to solve this problem, an improved private set intersection protocol based on Hadamard gate is proposed. Firstly, the more feasible Hadamard gates are used to perform on the original n qubits instead of the quantum Fourier transform, which may reduce the difficulty of implementation. In addition, through the exclusive OR calculation, the participant's private information is randomly chosen and encoded on the additional n qubits, which prevents participants from obtaining the result of the difference set S-diff , and then avoids the internal leakage of private information. Finally, the correctness and security analysis are conducted to show the proposed protocol can guarantee the correctness of computation result as well as resist outside attacks and participant internal attacks.

quant-ph↗
arXiv2023

Quantum Searchable Encryption for Cloud Data Based on Full-Blind Quantum Computation

Searchable encryption (SE) is a positive way to protect users sensitive data in cloud computing setting, while preserving search ability on the server side, i.e., it allows the server to search encrypted data without leaking information about the plaintext data. In this paper, a multi-client universal circuit-based full-blind quantum computation (FBQC) model is proposed. In order to meet the requirements of multi-client accessing or computing encrypted cloud data, all clients with limited quantum ability outsource the key generation to a trusted key center and upload their encrypted data to the data center. Considering the feasibility of physical implementation, all quantum gates in the circuit are replaced with the combination of π/8 rotation operator set {Rz(π/4), Ry(π/4), CRz(π/4), CRy(π/4), CCRz(π/4), CCRy(π/4)}. In addition, the data center is only allowed to perform one π/8 rotation operator each time, but does not know the structure of the circuit (i.e., quantum computation), so it can guarantee the blindness of computation. Then, through combining this multi-client FBQC model and Grover searching algorithm, we continue to propose a quantum searchable encryption scheme for cloud data. It solves the problem of multi-client access mode under searchable encryption in the cloud environment, and has the ability to resist against some quantum attacks. To better demonstrate our scheme, an example of our scheme to search on encrypted 2-qubit state is given in detail. Furthermore, the security of our scheme is analysed from two aspects: external attacks and internal attacks, and the result indicates that it can resist against such kinds of attacks and also guarantee the blindness of data and computation.

cs.ET↗
arXiv2023

PreNAS: Preferred One-Shot Learning Towards Efficient Neural Architecture Search

The wide application of pre-trained models is driving the trend of once-for-all training in one-shot neural architecture search (NAS). However, training within a huge sample space damages the performance of individual subnets and requires much computation to search for an optimal model. In this paper, we present PreNAS, a search-free NAS approach that accentuates target models in one-shot training. Specifically, the sample space is dramatically reduced in advance by a zero-cost selector, and weight-sharing one-shot training is performed on the preferred architectures to alleviate update conflicts. Extensive experiments have demonstrated that PreNAS consistently outperforms state-of-the-art one-shot NAS competitors for both Vision Transformer and convolutional architectures, and importantly, enables instant specialization with zero search cost. Our code is available at https://github.com/tinyvision/PreNAS.

cs.NI↗
arXiv2022

Balanced control between performance and saturation for constrained nonlinear systems

This paper addresses the balanced control between performance and saturation for a class of constrained nonlinear systems, including the branches: balanced command filtered backstepping (BCFB) and balanced performance control (BPC). To balance the interconnection and conflict between performance and saturation constraints, define a performance safety evaluation (PSE) function, which evaluates the system safety under the destabilizing effect variables (DEVs) like saturation quantity and filter errors, then the cumulative effects of DEVs are fully utilized and compensated for the performance recovery. Specifically, there exists some degree of tolerance for the DEVs in the safety region, and the compensation operation works when the evaluation of the system goes dangerous. The advantages of the proposed methodology are illustrated in the numerical simulation.

eess.SY↗
arXiv2022

HIE-SQL: History Information Enhanced Network for Context-Dependent Text-to-SQL Semantic Parsing

Recently, context-dependent text-to-SQL semantic parsing which translates natural language into SQL in an interaction process has attracted a lot of attention. Previous works leverage context-dependence information either from interaction history utterances or the previous predicted SQL queries but fail in taking advantage of both since of the mismatch between natural language and logic-form SQL. In this work, we propose a History Information Enhanced text-to-SQL model (HIE-SQL) to exploit context-dependence information from both history utterances and the last predicted SQL query. In view of the mismatch, we treat natural language and SQL as two modalities and propose a bimodal pre-trained model to bridge the gap between them. Besides, we design a schema-linking graph to enhance connections from utterances and the SQL query to the database schema. We show our history information enhanced methods improve the performance of HIE-SQL by a significant margin, which achieves new state-of-the-art results on the two context-dependent text-to-SQL benchmarks, the SparC and CoSQL datasets, at the writing time.

cs.DB↗
arXiv2022

Deformation mechanism of WC single crystals under nanoindentation: Effects of surface defects and orientation on pop-in and hysteresis

Nanoindentation was carried out on pure tungsten carbide (WC) on the basal (0001) and prismatic (1010) planes, using Berkovich and spherical indenters, in both single load and multi-load testing. The work focuses on correlating the load-displacement curves, including elastic to plastic deformation, size effect and hysteresis with the deformation behaviour of WC. With different specimen preparation processes, the elastic to plastic deformation started at different threshold loads: This observation was found to be due to the variation in surface dislocation density. Staircase deformation was observed thought to be caused by dislocation motion and the formation of slip bands; sudden displacement discontinuities in the load-displacement response - associated with dislocation loop nucleation - occurred at, or near the theoretical shear strength. Furthermore, discontinuities in load-displacement curves were also used to confirm that hysteresis loops were a result of plastic deformation, as they when the loading was purely elastic.

cond-mat.mtrl-sci↗
arXiv2021

The sharp lower bound of the first Dirichlet eigenvalue for geodesic balls

On complete noncompact Riemannian manifolds with non-negative Ricci curvature, Li-Schoen proved the uniform Poincare inequality for any ge odesic ball. In this note, we obtain the sharp lower bound of the first Dirichlet eigenvalue of such geodesic balls, which implies the sharp Poincare inequality for geodesic balls.

math.DG↗
arXiv2021

Sequential Mechanisms for Multi-type Resource Allocation

Several resource allocation problems involve multiple types of resources, with a different agency being responsible for "locally" allocating the resources of each type, while a central planner wishes to provide a guarantee on the properties of the final allocation given agents' preferences. We study the relationship between properties of the local mechanisms, each responsible for assigning all of the resources of a designated type, and the properties of a sequential mechanism which is composed of these local mechanisms, one for each type, applied sequentially, under lexicographic preferences, a well studied model of preferences over multiple types of resources in artificial intelligence and economics. We show that when preferences are O-legal, meaning that agents share a common importance order on the types, sequential mechanisms satisfy the desirable properties of anonymity, neutrality, non-bossiness, or Pareto-optimality if and only if every local mechanism also satisfies the same property, and they are applied sequentially according to the order O. Our main results are that under O-legal lexicographic preferences, every mechanism satisfying strategyproofness and a combination of these properties must be a sequential composition of local mechanisms that are also strategyproof, and satisfy the same combinations of properties.

cs.GT↗
arXiv2020

Multi-type Resource Allocation with Partial Preferences

We propose multi-type probabilistic serial (MPS) and multi-type random priority (MRP) as extensions of the well known PS and RP mechanisms to the multi-type resource allocation problem (MTRA) with partial preferences. In our setting, there are multiple types of divisible items, and a group of agents who have partial order preferences over bundles consisting of one item of each type. We show that for the unrestricted domain of partial order preferences, no mechanism satisfies both sd-efficiency and sd-envy-freeness. Notwithstanding this impossibility result, our main message is positive: When agents' preferences are represented by acyclic CP-nets, MPS satisfies sd-efficiency, sd-envy-freeness, ordinal fairness, and upper invariance, while MRP satisfies ex-post-efficiency, sd-strategy-proofness, and upper invariance, recovering the properties of PS and RP.

cs.AI↗
arXiv2020

Stability and collisions of quantum droplets in PT -symmetric dual-core couplers

We study the effect of the interplay between parity-time ($\mathcal{PT}$) symmetry and optical lattice (OL) potential on dynamics of quantum droplets (QDs) forming in a binary bosonic condensate trapped in a dual-core system. It is found that the stability of symmetric QDs in such non-Hermitian system depends critically on the competition of gain and loss $γ$, inter-core coupling $κ$, and OL potential. In the absence of OL potential, the $\mathcal{PT}$-symmetric QDs are unstable against symmetry-breaking perturbations with the increase of the total condensate norm $N$, and they retrieve the stability at larger $N$, in the weakly-coupled regime. As expected, the stable region of the $\mathcal{PT}$-symmetric QDs shrinks when $γ$ increases, i.e., the $\mathcal{PT}$ symmetry is prone to break the stability of QDs. There is a critical value of $κ$ beyond which the $\mathcal{PT}$-symmetric QDs are entirely stable in the unbroken $\mathcal{PT}$-symmetric phase. In the presence of OL potential, the $\mathcal{PT}$-symmetric on-site QDs are still stable for relatively small and large values of $N$. Nevertheless, it is demonstrated that the OL potential can assist stabilization of $\mathcal{PT}$-symmetric on-site QDs for some moderate values of $N$. On the other hand, it is worth noting that the relatively small $\mathcal{PT}$-symmetric off-site QDs are unstable, and only the relatively large ones are stable. Furthermore, collisions between stable $\mathcal{PT}$-symmetric QDs are considered too. It is revealed that the slowly moving $\mathcal{PT}$-symmetric QDs tend to merge into breathers, while the fast-moving ones display quasi-elastic collision and suffer fragmentation for small and large values of $N$, respectively.

nlin.PS↗