Searcharxiv⌕ Search

arXiv subjects

Xiangyu Wang

Publications and source records attributed to Xiangyu Wang.

At least 19 recordsLinked to original sources

SIMQ: A Single Graph-based Index for Multi-Vector Query

Multi-vector retrieval models such as ColBERT achieve state-of-the-art effectiveness by representing documents as sets of token-level vectors and scoring via MaxSim, but their high computational cost limits scalability to large corpora. Existing approximate methods face fundamental bottlenecks: centroid interaction matrix approaches require computing a full matrix at every query and rely on GPU-accelerated clustering, while data-oblivious compression methods yield unstable recall that requires repeated compensation. We propose SIMQ (Single graph Index for Multi-vector Query), a framework that bridges single-vector graph indices and multi-vector retrieval through data-aware sampling-based clustering and per-token graph navigation, simultaneously bypassing all aforementioned bottlenecks. SIMQ organizes all token vectors into fine-grained clusters via data-aware sampling, builds a single graph index over cluster centroids independently of PLAID's residual representation, and approximates the MaxSim score through per-token graph search without materializing any interaction matrix or relying on GPU acceleration. We provide theoretical analysis showing that, under a manifold assumption and successful cell retrieval, data-aware partitioning has a smaller expected approximation error than a data-oblivious partition. Building on this foundation, SIMQ+ further introduces count-based scoring and a dynamic candidate set to improve robustness across varying query conditions. Experiments on five benchmarks compare recall, throughput, and construction time against several baselines, primarily on CPU; large-scale PLAID k-means centroids are built on a separate GPU machine. SIMQ has been integrated into VSAG, an Ant Group's open-source vector search library.

cs.IR↗

On Oscillatory Integral Operators Satisfying the cinematic curvature condition

We establish improved $L^r\mapsto L^p$ estimates for oscillatory integral operators satisfying the positive cinematic curvature condition. In ambient dimension $n\geq4$, the output range is $p>2+8/(3n-4)$, with the full Knapp-admissible input range $r'\leq (n-2)p/n$. The proof combines a transverse covering $L^2$ estimate, Wolff's two-ends reduction, and refined decoupling. In dimension $n=3$ we retain the range $p>4$.

math.CA↗

Wave Packet Density in Local Smoothing: Refinements and Limitations

We develop a transverse covering $L^2$ argument that both substantially simplifies the approach of Gan, He, Li, and Wu and improves their local smoothing estimates for Fourier integral operators with positive conic curvature in even spatial dimensions $n\geq 4$. We also investigate the limitations of wave packet density methods, showing that genuine geometric obstructions already arise in the Euclidean setting.

math.CA↗

An optical-frequency-comb-enabled unified infrastructure for quantum networks

Quantum networks require precision timing and classical communication to coordinate distributed quantum operations, yet these functions are typically supported by separate infrastructures, increasing system complexity and limiting scalability. Here we demonstrate an optical-frequency-comb-enabled infrastructure that unifies precision timing, classical communication and quantum communication over a shared fibre link. By encoding classical data onto the intensity of the comb while retaining its pulse timing, we simultaneously transmit data at 100 Mbit/s and achieve a timing stability of 0.85 fs at 1 s over 100 km. We develop a spectral-engineering strategy to suppress the coupling between data modulation and timing extraction. We further establish a power-dependent model of the excess noise induced in a co-propagating continuous-variable quantum key distribution channel, identifying optical power, wavelength allocation and propagation direction as the principal control parameters. After optimization, the system introduces only 0.0037 shot noise unit of additional excess noise and limits the secret-key-rate penalty to 6.7%. Furthermore, quantum-secured timestamp transmission additionally enables absolute time synchronization with a stability of 0.26 fs at 1 s. These results establish optical frequency combs as a common physical layer for integrated quantum-network services.

physics.optics↗

UAV-Flow++: A Unified Suite for Benchmarking and Learning Language-Guided UAV Flight

Language-guided UAV navigation has largely focused on long-horizon route planning, leaving short-range, object-relative flight maneuvers underexplored. To address this gap, we formulate Flying-on-a-Word (Flow), a task requiring a UAV to convert concise natural-language instruction, egocentric visual observation, and current state dynamics into continuous flight actions. Flow encompasses both primitive motion commands and complex object-interactive behaviors, assessing motion intent comprehension and spatial context grounding. To enable policy learning and closed-loop evaluation, we introduce UAV-Flow, a comprehensive dataset comprising 30,692 expert-piloted real-world trajectories across eight distinct motion categories, alongside 10,109 simulation trajectories in the UAV-Flow-Sim environment equipped with task-specific completion criteria. Furthermore, we develop FlowVLA, which undergoes imitation pretraining on mixed real-sim demonstrations and online post-training via Tail-Aware GRPO (TGRPO). By leveraging group reward distributions, TGRPO effectively reinforces high-reward action sequences. Evaluated on 273 held-out UAV-Flow-Sim episodes, pretrained FlowVLA achieves a mean success rate of 75.58%, outperforming the strongest baseline by 6.72 percentage points, which TGRPO post-training further elevates to 93.16% with superior sample efficiency. Physical UAV experiments demonstrate robust execution of object-relative commands in both indoor and outdoor real-world scenes. Together, Flow establishes a novel benchmark and learning framework for language-conditioned flight maneuvers.

cs.RO↗

Reversing Heat Flow by Coherence in a Multipartite Quantum System

The second law of thermodynamics dictates that heat flows spontaneously from a high-temperature entity to a lower-temperature one. Yet, recent advances have demonstrated that quantum correlations between a system and its thermal environment can induce a reversal of heat flow, challenging classical thermodynamic expectations. Here, we experimentally demonstrate that internal quantum coherence in a multipartite spin system can also reverse heat flow, without relying on correlations with the environment. Under the collision model with cascade interaction, we verify that both the strength and the phase of the coherence term determine the direction and magnitude of energy transfer.These results enable precise control of heat flow using only local quantum properties.

quant-ph↗

Cone Restriction Estimates under Half-Scale Dimensional Conditions

We prove cone restriction estimates under half-scale dimensional conditions. Building on the framework of Du--Li--Wang--Zhang, we incorporate the broad restriction estimate of Ou--Wang and a localized multilinear estimate obtained through cyclic bilinear pairing and multilinear Kakeya bounds for local packet energies. For certain ambient and weight dimensions, our estimates are sharp in suitable ranges of $p$, up to an $R^\varepsilon$ loss.

math.CA↗

AVSG: Accelerated Vectorized Sparse Gather for Efficient KV Cache Offload in Sparse-Attention LLM Serving

Dynamic sparse attention reduces long-context attention computation by selecting only a subset of tokens, but still requires access to the full KV cache, leaving serving memory-bound. Offloading the KV cache to host memory reduces device memory pressure but places H2D transfers on the decoding critical path. In DSA, substantial overlap in selected KV entries across decoding steps creates an opportunity for device-resident reuse. Exploiting this reuse efficiently, however, presents three critical challenges: costly matching of selected tokens against entries retained in the HBM buffer, uneven distribution of the remaining H2D transfers across accelerator cores, and retaining frequently accessed KV entries within limited HBM capacity. We present AVSG, an operator that addresses these challenges for DSA while preserving its exact token selections. AVSG uses vectorized hash matching to identify reusable HBM-buffer slots and reserve slots for entries requiring transfer, then evenly partitions these H2D transfers across accelerator cores. Lifetime-based buffer management retains frequently accessed KV entries in the device buffer across decoding steps, and shared slot reservations extend reuse across tokens within a multi-token prediction iteration. On a single NPU, vectorized hash matching is 2.80x faster than scalar dual-pointer matching, miss-only transfer raises effective H2D bandwidth by up to 34.44x over request-level assignment, and an 8K-entry buffer reaches a 94.83% hit rate on real requests. These gains translate into end-to-end improvement: on a serving stack processing a real-world production dataset, AVSG reduces time per output token by 39% and increases output throughput by 1.27x relative to the same KV offload layout without HBM-buffer reuse. The code of AVSG is released at https://gitcode.com/cann/cann-recipes-infer.

eess.SP↗

SuiChat-CN: Benchmarking Contextual Suicide Risk Assessment in Chinese Group Chats

Suicide is a critical global public health challenge, causing approximately 720,000 deaths each year and calling for timely, effective prevention strategies. Existing computational studies primarily focus on post-based social media platforms such as Twitter and Weibo, leaving instant messaging environments such as Telegram underexplored. Yet group chats pose distinct challenges: messages are short, fragmented, multi-party, and often rely on implicit or culturally specific expressions, making isolated post-level analysis insufficient. We introduce SuiChat-CN, a Chinese group-chat benchmark for contextual suicide risk assessment. We collect public Telegram group-chat data, construct coherent conversational segments through signal-word extraction and bidirectional context expansion, and annotate user risk levels with an expert-validated, LLM-assisted paradigm. SuiChat-CN contains 13,312 contextual segments from 1,406 users, covering 258,228 raw chat messages. Extensive experiments with PLMs and more than 40 LLMs demonstrate that contextual information is essential for reliable risk assessment, while fine-tuning and partial-context evaluation further reveal the challenges of early detection in multi-party conversations. Due to ethical and sensitivity concerns, the dataset is not publicly released but will be shared with accredited mental health and suicide-prevention research institutions upon reasonable request.

cs.AI↗

FlowCritic: Bridging Value Estimation with Flow Matching in Reinforcement Learning

Reliable value estimation serves as the cornerstone of reinforcement learning (RL) by evaluating long-term returns and guiding policy improvement, significantly influencing the convergence speed and final performance. Existing works improve the reliability of value function estimation via multi-critic ensembles and distributional RL, yet the former merely combines multi point estimation without capturing distributional information, whereas the latter relies on discretization or quantile regression, limiting the expressiveness of complex value distributions. Inspired by flow matching's success in generative modeling, we propose a generative paradigm for value estimation, named FlowCritic. Departing from conventional regression for deterministic value prediction, FlowCritic leverages flow matching to model value distributions and generate samples for value estimation.

cs.LG↗

Restriction estimates using multilinear Kakeya

We use multilinear Kakeya estimates to establish restriction estimates in higher dimensions. For sufficiently large dimensions, our approach yields better exponents than those obtained by a direct application of polynomial partitioning.

math.CA↗

L2G-SR: Learning to Guide LLM-based Symbolic Regression

Symbolic regression searches for an equation that explains observed data. Numerical error is commonly used to guide the search, but it alone can favor an accurate surrogate over a candidate with the right structure. We propose L2G-SR, a framework for LLM-based symbolic regression that guides search with a learned structural verifier. We construct structural supervision signal using tree edit distance (TED), accounting for commutative operators and alternative expression forms. A verifier first aligns numerical observations with the formulas, then learns to score a candidate's structural similarity to the target formula. At search time, the target formula is hidden and the verifier complements numerical error in parent selection. An analyst then uses verifier feedback and diagnostic tools to propose revisions, which a generator converts into refined expressions. On LLM-SRBench, L2G-SR achieves state-of-the-art numerical fit, equation level accuracy and TED across the evaluated LLM backbones and domains. L2G-SR demonstrates the value of learned structural guidance for LLM-based symbolic regression.

cs.AI↗

A Short Proof of a Theorem of Pan and Zhang

Pan and Zhang proved that a positive proportion of positive integers can be represented in the form $a^2+b^2+2^{c^2}$, where $(a,b,c)$ are nonnegative integers. In this note, we give a short proof of their result. The proof uses Romanoff's method and an elementary divisor argument to control the relevant second moment.

math.NT↗

On integers whose friable component is large

We estimate the number of integers up to $x$ whose $y$-friable component exceeds $z$. Uniformly for $\max\{2,\log z\}\leqslant y\leqslant z\leqslant x/2$, we determine the logarithm of this counting function up to an error $O(\log z/\log y)$. This sharpens an estimate arising from a problem in Ford's sieve methods lecture notes. The proof combines classical friable-number estimates with an elementary argument for the lower bound.

math.NT↗

STLQ: Spatio-Topological Linkage Quantization for Approximate Nearest Neighbor Search

Vector quantization is widely used for large-scale data compression and approximate nearest neighbor search (ANNS). Multi-codebook quantization (MCQ) is a practical paradigm that represents database vectors with compact codes and supports efficient retrieval through asymmetric distance computation (ADC). However, existing shallow MCQs encode vectors independently within each inverted file (IVF) cluster, without exploiting local references. Therefore, we propose Spatio-Topological Linkage Quantization (STLQ), a topological shallow MCQ method that organizes each IVF cluster as a directed-forest with compact codes and scaling coefficients. Its basic encoding phase first quantizes vectors independently, while the linked encoding phase re-encodes relative residuals with respect to selected reconstructed parents, reusing local reconstruction information to obtain finer representations. The corresponding linked-ADC preserves linear scanning complexity. STLQ further compresses the scaling coefficients and directed-forest structure to control the additional topology overhead. Extensive experiments show that STLQ achieves higher recall than the evaluated shallow MCQs at comparable bit budgets, although at a higher offline construction cost. Meanwhile, it has a competitive recall-QPS tradeoff against deep MCQs. The code is available at https://github.com/ighxiy/STLQ.

cs.CV↗

A weighted restriction estimate in $\mathbb R^2$

We establish a weighted restriction estimate in $\mathbb{R}^2$. Our proof combines new two-ends Furstenberg estimates with refined decoupling.As an application, our result yields decay estimates for the circular $L^p$-means of the Fourier transforms of fractal measures in $\mathbb{R}^2$.

math.CA↗

On Cone Restriction Estimates in Higher Dimensions

We revisit the Ou-Wang's approach to the cone restriction problem via polynomial partitioning. By recasting their inductive scheme as a recursive algorithm and incorporating the nested polynomial Wolff axioms, we obtain improved bounds for cone restriction estimates in higher dimensions.

math.CA↗

AnyStep-WAM: Budget-Aligned Distillation and Adaptive Inference for World Action Models

World-action models (WAMs) couple predictive visual modeling with action generation, typically relying on iterative denoising with a fixed denoising steps. However, manipulation tasks contain actions chunks with varying sensitivity to generation errors: critical actions require precision, while less sensitive actions allow faster generation with fewer denoising steps. Here we introduce AnyStep World Action Model, a general framework for tunable-budget prediction and scene-dependent computation allocation. Our budget-aligned teacher-trajectory distillation trains interval-conditioned flow maps using explicit frozen-teacher transitions and shared low-rank adapters, supporting action generation from one-step prediction to multi-step refinement. Building on this capability, a lightweight risk-benefit scheduler predicts teacher-curvature-based difficulty and budget-specific student-teacher fidelity from a single one-step preview, selecting the smallest budget predicted to satisfy risk-adaptive fidelity requirements. We evaluate our framework on three widely used WAMs Motus, FastWAM, and LingBotVA using RoboTwin 2.0. Our method reduces average denoising steps by 60.2%, 49.8%, and 85.28%, respectively, while maintaining baseline task success rates. In particular, our AnyStep training substantially improves model performance under a one-step denoising budget, increasing task success rates by 7.07%, 12.08%, and 8.94% on Motus, FastWAM, and LingBotVA, respectively. Experiments on six real-world manipulation tasks further validate its effectiveness.

cs.RO↗