arXiv subjects

Qinwen Wang

Publications and source records attributed to Qinwen Wang.

At least 19 recordsLinked to original sources

arXiv2026

GateDiffInt: Gate-Mediated Controllable Diffusion and Multi-Intent LLM Distillation for User Behavior Modeling

Existing ranking models encode intent only implicitly, making it hard to disentangle structured intents of varying strength and temporal scale. Noise and intent in behavior sequences are mutually reinforcing---we call this Noise--Intent Coupling (NIC). Noise dilutes true intents, while the lack of structured intent priors leaves denoising without a clear target. To address NIC, we propose GateDiffInt, an intent interaction framework for industrial ranking. It uses the final conversion signal to jointly align sequence denoising and intent extraction. GateDiffInt applies a controllable forward diffusion process with dual gating to enhance and denoise behavior sequences. A large language model then acts as teacher to distill four structured intents---long-term, short-term, latent, and conversion into a lightweight student model. The enhanced sequence and structured intent representations are deeply fused via attention to produce intent-aware representations for conversion-rate prediction. Extensive experiments on public and large-scale industrial datasets show consistent gains over strong baselines. In online A/B tests serving hundreds of millions of daily active users, GateDiffInt delivers substantial GMV improvements and has been deployed to primary traffic, confirming both effectiveness and production readiness.

cs.IR↗

arXiv2026

Mamba with Hierarchical Memory: Solving Representation Bottleneck in Long Sequence Modeling

Recurrent linear attention models (RLAs) such as Mamba offer efficient linear-time sequence modeling as an alternative to Transformers, yet their fixed-capacity recurrent states limit long-sequence modeling. Drawing inspiration from hierarchical human memory, we propose Hierarchical Memory Mamba (HMM) to address this limitation. Building upon a pre-trained Mamba backbone, HMM integrates a lightweight working memory that extracts slow paragraph-level semantics (PLS) from the fast sensory memory embedded in the backbone's hidden states. The PLS is subsequently compressed into persistent long-term memory for task-relevant retrieval. The hierarchical processing of semantic information overcomes the representation bottleneck of RLAs and endows HMM cross-task generalization through parametric learning, which is not observed in other long-context enhanced Mamba variants. Evaluations on Passkey Retrieval and LongBench-E tasks demonstrate that HMM improves retrieval success by 34.3--37.1% and reasoning accuracy by 1.6--14.2% over strong Mamba-based models, while adding only 2% extra parameters and with minimal training overhead.

cs.AI↗

arXiv2026

PiCA: Pivot-Based Credit Assignment for Search Agentic Reinforcement Learning

Large Language Model (LLM)-based search agents trained with reinforcement learning (RL) have significantly improved the performance of knowledge-intensive tasks. However, existing methods encounter critical challenges in long-horizon credit assignment: (i) Reward Sparsity, where models receive only outcome feedback without step-level guidance to differentiate action quality; (ii) Isolated Credit, where credit is assigned to steps independently, failing to capture sequential dependencies; and (iii) Distributional Shift, where rewards are estimated on templates that deviate from the model's natural generative distribution. To address these issues, we propose Pivot-Based Credit Assignment (PiCA), a novel step reward mechanism that reformulates the search trajectory as a sequential process of cumulative search progress. Unlike prior isolated step rewards, PiCA defines process rewards as success probabilities dependent on the historical context based on Potential-Based Reward Shaping (PBRS). This approach identifies pivot steps, which comprise target golden sub-queries and sub-answers derived from historical trajectories, as information peaks that significantly boost the likelihood of a correct final answer. By anchoring these step rewards to the final task objective, PiCA provides dense, pivot-aware and trajectory-dependent guidance while maintaining distributional consistency. Extensive experiments show that PiCA outperforms existing strong baselines across seven knowledge-intensive QA benchmarks, achieving 15.2% and 2.2% improvements for 3B and 7B models. The consistent performance gains across various models show PiCA's robust generalization. The code is available at https://github.com/novdream/PiCA.

cs.AI↗

arXiv2023

The NeurIPS 2022 Neural MMO Challenge: A Massively Multiagent Competition with Specialization and Trade

In this paper, we present the results of the NeurIPS-2022 Neural MMO Challenge, which attracted 500 participants and received over 1,600 submissions. Like the previous IJCAI-2022 Neural MMO Challenge, it involved agents from 16 populations surviving in procedurally generated worlds by collecting resources and defeating opponents. This year's competition runs on the latest v1.6 Neural MMO, which introduces new equipment, combat, trading, and a better scoring system. These elements combine to pose additional robustness and generalization challenges not present in previous competitions. This paper summarizes the design and results of the challenge, explores the potential of this environment as a benchmark for learning methods, and presents some practical reinforcement learning training approaches for complex tasks with sparse rewards. Additionally, we have open-sourced our baselines, including environment wrappers, benchmarks, and visualization tools for future research.

cs.AI↗

arXiv2023

Progression Cognition Reinforcement Learning with Prioritized Experience for Multi-Vehicle Pursuit

Multi-vehicle pursuit (MVP) such as autonomous police vehicles pursuing suspects is important but very challenging due to its mission and safety critical nature. While multi-agent reinforcement learning (MARL) algorithms have been proposed for MVP problem in structured grid-pattern roads, the existing algorithms use randomly training samples in centralized learning, which leads to homogeneous agents showing low collaboration performance. For the more challenging problem of pursuing multiple evading vehicles, these algorithms typically select a fixed target evading vehicle for pursuing vehicles without considering dynamic traffic situation, which significantly reduces pursuing success rate. To address the above problems, this paper proposes a Progression Cognition Reinforcement Learning with Prioritized Experience for MVP (PEPCRL-MVP) in urban multi-intersection dynamic traffic scenes. PEPCRL-MVP uses a prioritization network to assess the transitions in the global experience replay buffer according to the parameters of each MARL agent. With the personalized and prioritized experience set selected via the prioritization network, diversity is introduced to the learning process of MARL, which can improve collaboration and task related performance. Furthermore, PEPCRL-MVP employs an attention module to extract critical features from complex urban traffic environments. These features are used to develop progression cognition method to adaptively group pursuing vehicles. Each group efficiently target one evading vehicle in dynamic driving environments. Extensive experiments conducted with a simulator over unstructured roads of an urban area show that PEPCRL-MVP is superior to other state-of-the-art methods. Specifically, PEPCRL-MVP improves pursuing efficiency by 3.95% over TD3-DMAP and its success rate is 34.78% higher than that of MADDPG. Codes are open sourced.

cs.AI↗

arXiv2022

An Opponent-Aware Reinforcement Learning Method for Team-to-Team Multi-Vehicle Pursuit via Maximizing Mutual Information Indicator

The pursuit-evasion game in Smart City brings a profound impact on the Multi-vehicle Pursuit (MVP) problem, when police cars cooperatively pursue suspected vehicles. Existing studies on the MVP problems tend to set evading vehicles to move randomly or in a fixed prescribed route. The opponent modeling method has proven considerable promise in tackling the non-stationary caused by the adversary agent. However, most of them focus on two-player competitive games and easy scenarios without the interference of environments. This paper considers a Team-to-Team Multi-vehicle Pursuit (T2TMVP) problem in the complicated urban traffic scene where the evading vehicles adopt the pre-trained dynamic strategies to execute decisions intelligently. To solve this problem, we propose an opponent-aware reinforcement learning via maximizing mutual information indicator (OARLM2I2) method to improve pursuit efficiency in the complicated environment. First, a sequential encoding-based opponents joint strategy modeling (SEOJSM) mechanism is proposed to generate evading vehicles' joint strategy model, which assists the multi-agent decision-making process based on deep Q-network (DQN). Then, we design a mutual information-united loss, simultaneously considering the reward fed back from the environment and the effectiveness of opponents' joint strategy model, to update pursuing vehicles' decision-making process. Extensive experiments based on SUMO demonstrate our method outperforms other baselines by 21.48% on average in reducing pursuit time. The code is available at \url{https://github.com/ANT-ITS/OARLM2I2}.

cs.MA↗

arXiv2022

Graded-Q Reinforcement Learning with Information-Enhanced State Encoder for Hierarchical Collaborative Multi-Vehicle Pursuit

The multi-vehicle pursuit (MVP), as a problem abstracted from various real-world scenarios, is becoming a hot research topic in Intelligent Transportation System (ITS). The combination of Artificial Intelligence (AI) and connected vehicles has greatly promoted the research development of MVP. However, existing works on MVP pay little attention to the importance of information exchange and cooperation among pursuing vehicles under the complex urban traffic environment. This paper proposed a graded-Q reinforcement learning with information-enhanced state encoder (GQRL-IESE) framework to address this hierarchical collaborative multi-vehicle pursuit (HCMVP) problem. In the GQRL-IESE, a cooperative graded Q scheme is proposed to facilitate the decision-making of pursuing vehicles to improve pursuing efficiency. Each pursuing vehicle further uses a deep Q network (DQN) to make decisions based on its encoded state. A coordinated Q optimizing network adjusts the individual decisions based on the current environment traffic information to obtain the global optimal action set. In addition, an information-enhanced state encoder is designed to extract critical information from multiple perspectives and uses the attention mechanism to assist each pursuing vehicle in effectively determining the target. Extensive experimental results based on SUMO indicate that the total timestep of the proposed GQRL-IESE is less than other methods on average by 47.64%, which demonstrates the excellent pursuing efficiency of the GQRL-IESE. Codes are outsourced in https://github.com/ANT-ITS/GQRL-IESE.

cs.LG↗

arXiv2022

On eigenvalues of a high-dimensional Kendall's rank correlation matrix with dependence

This paper investigates limiting spectral distribution of a high-dimensional Kendall's rank correlation matrix. The underlying population is allowed to have general dependence structure. The result no longer follows the generalized Marucenko-Pastur law, which is brand new. It's the first result on rank correlation matrices with dependence. As applications, we study the Kendall's rank correlation matrix for multivariate normal distributions with a general covariance matrix. From these results, we further gain insights on Kendall's rank correlation matrix and its connections with the sample covariance/correlation matrix.

math.ST↗

arXiv2022

Confidence Estimation Transformer for Long-term Renewable Energy Forecasting in Reinforcement Learning-based Power Grid Dispatching

The expansion of renewable energy could help realizing the goals of peaking carbon dioxide emissions and carbon neutralization. Some existing grid dispatching methods integrating short-term renewable energy prediction and reinforcement learning (RL) have been proved to alleviate the adverse impact of energy fluctuations risk. However, these methods omit the long-term output prediction, which leads to stability and security problems on the optimal power flow. This paper proposes a confidence estimation Transformer for long-term renewable energy forecasting in reinforcement learning-based power grid dispatching (Conformer-RLpatching). Conformer-RLpatching predicts long-term active output of each renewable energy generator with an enhanced Transformer to boost the performance of hybrid energy grid dispatching. Furthermore, a confidence estimation method is proposed to reduce the prediction error of renewable energy. Meanwhile, a dispatching necessity evaluation mechanism is put forward to decide whether the active output of a generator needs to be adjusted. Experiments carried out on the SG-126 power grid simulator show that Conformer-RLpatching achieves great improvement over the second best algorithm DDPG in security score by 25.8% and achieves a better total reward compared with the golden medal team in the power grid dispatching competition sponsored by State Grid Corporation of China under the same simulation environment. Codes are outsourced in https://github.com/buptlxh/Conformer-RLpatching.

cs.LG↗

arXiv2022

$ \text{T}^3 $OMVP: A Transformer-based Time and Team Reinforcement Learning Scheme for Observation-constrained Multi-Vehicle Pursuit in Urban Area

Smart Internet of Vehicles (IoVs) combined with Artificial Intelligence (AI) will contribute to vehicle decision-making in the Intelligent Transportation System (ITS). Multi-Vehicle Pursuit games (MVP), a multi-vehicle cooperative ability to capture mobile targets, is becoming a hot research topic gradually. Although there are some achievements in the field of MVP in the open space environment, the urban area brings complicated road structures and restricted moving spaces as challenges to the resolution of MVP games. We define an Observation-constrained MVP (OMVP) problem in this paper and propose a Transformer-based Time and Team Reinforcement Learning scheme ($ \text{T}^3 $OMVP) to address the problem. First, a new multi-vehicle pursuit model is constructed based on decentralized partially observed Markov decision processes (Dec-POMDP) to instantiate this problem. Second, by introducing and modifying the transformer-based observation sequence, QMIX is redefined to adapt to the complicated road structure, restricted moving spaces and constrained observations, so as to control vehicles to pursue the target combining the vehicle's observations. Third, a multi-intersection urban environment is built to verify the proposed scheme. Extensive experimental results demonstrate that the proposed $ \text{T}^3 $OMVP scheme achieves significant improvements relative to state-of-the-art QMIX approaches by 9.66%~106.25%. Code is available at https://github.com/pipihaiziguai/T3OMVP.

cs.AI↗

arXiv2021

Eigenvalue distribution of a high-dimensional distance covariance matrix with application

We introduce a new random matrix model called distance covariance matrix in this paper, whose normalized trace is equivalent to the distance covariance. We first derive a deterministic limit for the eigenvalue distribution of the distance covariance matrix when the dimensions of the vectors and the sample size tend to infinity simultaneously. This limit is valid when the vectors are independent or weakly dependent through a finite-rank perturbation. It is also universal and independent of the details of the distributions of the vectors. Furthermore, the top eigenvalues of this distance covariance matrix are shown to obey an exact phase transition when the dependence of the vectors is of finite rank. This finding enables the construction of a new detector for such weak dependence where classical methods based on large sample covariance matrices or sample canonical correlations may fail in the considered high-dimensional framework.

math.ST↗

arXiv2021

On eigenvalues of a high-dimensional spatial-sign covariance matrix

This paper investigates limiting properties of eigenvalues of multivariate sample spatial-sign covariance matrices when both the number of variables and the sample size grow to infinity. The underlying p-variate populations are general enough to include the popular independent components model and the family of elliptical distributions. A first result of the paper establishes that the distribution of the eigenvalues converges to a deterministic limit that belongs to the family of generalized Marcenko-Pastur distributions. Furthermore, a new central limit theorem is established for a class of linear spectral statistics. We develop two applications of these results to robust statistics for a high-dimensional shape matrix. First, two statistics are proposed for testing the sphericity. Next, a spectrum-corrected estimator using the sample spatial-sign covariance matrix is proposed. Simulation experiments show that in high dimension, the sample spatial-sign covariance matrix provides a valid and robust tool for mitigating influence of outliers.

math.ST↗

arXiv2020

Provable More Data Hurt in High Dimensional Least Squares Estimator

This paper investigates the finite-sample prediction risk of the high-dimensional least squares estimator. We derive the central limit theorem for the prediction risk when both the sample size and the number of features tend to infinity. Furthermore, the finite-sample distribution and the confidence interval of the prediction risk are provided. Our theoretical results demonstrate the sample-wise nonmonotonicity of the prediction risk and confirm "more data hurt" phenomenon.

stat.ML↗

arXiv2019

Central Limit Theorem for Linear Spectral Statistics of Large Dimensional Kendall's Rank Correlation Matrices and its Applications

This paper is concerned with the limiting spectral behaviors of large dimensional Kendall's rank correlation matrices generated by samples with independent and continuous components. We do not require the components to be identically distributed, and do not need any moment conditions, which is very different from the assumptions imposed in the literature of random matrix theory. The statistical setting in this paper covers a wide range of highly skewed and heavy-tailed distributions. We establish the central limit theorem (CLT) for the linear spectral statistics of the Kendall's rank correlation matrices under the Marchenko-Pastur asymptotic regime, in which the dimension diverges to infinity proportionally with the sample size. We further propose three nonparametric procedures for high dimensional independent test and their limiting null distributions are derived by implementing this CLT. Our numerical comparisons demonstrate the robustness and superiority of our proposed test statistics under various mixed and heavy-tailed cases.

math.ST↗

arXiv2015

Identifying the number of factors from singular values of a large sample auto-covariance matrix

Identifying the number of factors in a high-dimensional factor model has attracted much attention in recent years and a general solution to the problem is still lacking. A promising ratio estimator based on the singular values of the lagged autocovariance matrix has been recently proposed in the literature and is shown to have a good performance under some specific assumption on the strength of the factors. Inspired by this ratio estimator and as a first main contribution, this paper proposes a complete theory of such sample singular values for both the factor part and the noise part under the large-dimensional scheme where the dimension and the sample size proportionally grow to infinity. In particular, we provide the exact description of the phase transition phenomenon that determines whether a factor is strong enough to be detected with the observed sample singular values. Based on these findings and as a second main contribution of the paper, we propose a new estimator of the number of factors which is strongly consistent for the detection of all significant factors (which are the only theoretically detectable ones). In particular, factors are assumed to have the minimum strength above the phase transition boundary which is of the order of a constant; they are thus not required to grow to infinity together with the dimension (as assumed in most of the existing papers on high-dimensional factor models). Empirical Monte-Carlo study as well as the analysis of stock returns data attest a very good performance of the proposed estimator. In all the tested cases, the new estimator largely outperforms the existing estimator using the same ratios of singular values.

stat.ME↗

arXiv2015

Extreme eigenvalues of large-dimensional spiked Fisher matrices with application

Consider two $p$-variate populations, not necessarily Gaussian, with covariance matrices $Σ_1$ and $Σ_2$, respectively, and let $S_1$ and $S_2$ be the sample covariances matrices from samples of the populations with degrees of freedom $T$ and $n$, respectively. When the difference $Δ$ between $Σ_1$ and $Σ_2$ is of small rank compared to $p,T$ and $n$, the Fisher matrix $F=S_2^{-1}S_1$ is called a {\em spiked Fisher matrix}. When $p,T$ and $n$ grow to infinity proportionally, we establish a phase transition for the extreme eigenvalues of $F$: when the eigenvalues of $Δ$ ({\em spikes}) are above (or under) a critical value, the associated extreme eigenvalues of the Fisher matrix will converge to some point outside the support of the global limit (LSD) of other eigenvalues; otherwise, they will converge to the edge points of the LSD. Furthermore, we derive central limit theorems for these extreme eigenvalues of the spiked Fisher matrix. The limiting distributions are found to be Gaussian if and only if the corresponding population spike eigenvalues in $Δ$ are {\em simple}. Numerical examples are provided to demonstrate the finite sample performance of the results. In addition to classical applications of a Fisher matrix in high-dimensional data analysis, we propose a new method for the detection of signals allowing an arbitrary covariance structure of the noise. Simulation experiments are conducted to illustrate the performance of this detector.

math.ST↗

arXiv2015

On singular values distribution of a large auto-covariance matrix in the ultra-dimensional regime

Let $(\varepsilon_{t})_{t>0}$ be a sequence of independent real random vectors of $p$-dimension and let $X_T=\sum_{t=s+1}^{s+T}\varepsilon_t\varepsilon^T_{t-s}/T$ be the lag-$s$ ($s$ is a fixed positive integer) auto-covariance matrix of $\varepsilon_t$. This paper investigates the limiting behavior of the singular values of $X_T$ under the so-called {\em ultra-dimensional regime} where $p\to\infty$ and $T\to\infty$ in a related way such that $p/T\to 0$. First, we show that the singular value distribution of $X_T$ after a suitable normalization converges to a nonrandom limit $G$ (quarter law) under the forth-moment condition. Second, we establish the convergence of its largest singular value to the right edge of $G$. Both results are derived using the moment method.

math.PR↗

arXiv2014

Joint CLT for several random sesquilinear forms with applications to large-dimensional spiked population models

In this paper, we derive a joint central limit theorem for random vector whose components are function of random sesquilinear forms. This result is a natural extension of the existing central limit theory on random quadratic forms. We also provide applications in random matrix theory related to large-dimensional spiked population models. For the first application, we find the joint distribution of grouped extreme sample eigenvalues correspond to the spikes. And for the second application, under the assumption that the population covariance matrix is diagonal with $k$ (fixed) simple spikes, we derive the asymptotic joint distribution of the extreme sample eigenvalue and its corresponding sample eigenvector projection.

math.PR↗