arXiv subjects

Hongli Xu

Publications and source records attributed to Hongli Xu.

At least 37 records · Page 2Linked to original sources

arXiv2026

Enabling Reconfiguration-Communication Overlap for Collective Communication in Optical Networks

Collective communication (CC) is critical for scaling distributed machine learning (DML). The predictable traffic patterns of DML present a great opportunity for applying optical network technologies. Optical networks with reconfigurable topologies promise high bandwidth and low latency for collective communications. However, existing approaches face inherent limitations: static topologies are inefficient for dynamic communication patterns within CC algorithm, while frequent topology reconfiguration matching every step of the algorithm incurs significant overhead. In this paper, we propose SWOT, a demand-aware optical network framework that employs ``intra-collective reconfiguration'' to dynamically align network resources with CC traffic patterns. SWOT hides reconfiguration latency by overlapping it with data transmission through three key techniques: \textit{Heterogeneous Message Splitting}, \textit{Asynchronous Overlapping}, and \textit{Topology Bypassing}. Extensive simulations demonstrate that SWOT reduces communication completion time up to 89.7% across diverse CC algorithm compared to static baselines, demonstrating strong robustness to varying optical resources and reconfiguration delay.

cs.NI↗

arXiv2026

Adaptive and Fine-grained Module-wise Expert Pruning for Efficient LoRA-MoE Fine-Tuning

LoRA-MoE has emerged as an effective paradigm for parameter-efficient fine-tuning, combining the low training cost of LoRA with the increased adaptation capacity of Mixture-of-Experts (MoE). However, existing LoRA-MoE frameworks typically adopt a fixed and uniform expert configuration across heterogeneous Transformer modules (\eg, attention query/key projections and MLP gating networks), ignoring their distinct functional roles and capacity requirements. This design leads to localized over-provisioning, redundant trainable parameters, and unnecessary optimizer-state overhead. Moreover, prior methods enforce load balancing among experts throughout training. Although beneficial in the early stage, this constraint becomes restrictive once routing patterns stabilize, limiting expert specialization on downstream tasks. In this paper, we propose DMEP, a novel LoRA-MoE fine-tuning framework based on Dynamic Module-wise Expert Pruning. DMEP tracks expert utilization during training and physically removes low-utility experts on a per-module basis, yielding a more compact expert structure tailored to different modules. The pruned model then continues training without the load-balancing constraint, freeing the remaining experts to focus entirely on the downstream task and develop specialized expertise. By jointly adapting module-wise expert capacity and eliminating unnecessary balancing, DMEP improves both parameter efficiency and training efficiency. Extensive experiments on multiple reasoning benchmarks show that DMEP reduces trainable parameters by 35\%--43\% and improves training throughput by about 10\%, while maintaining or surpassing the downstream reasoning accuracy of uniform LoRA-MoE baselines.

cs.LG↗

arXiv2026

Less is More: Lightweight Prompt Compression for Question Answering Applications on Edge Devices

In agent-driven question answering (QA) applications, retrieval-augmented generation (RAG) is commonly introduced to enhance the response accuracy of large language models (LLMs) by providing additional context. Due to the inherent noise in retrieval results and the coarse granularity of document-level retrieval, the retrieved context often contains substantial redundant information. In this setting, the agent prompt, consisting of the user query and the associated retrieved context, leads to unnecessary computational overhead during LLM inference. Existing prompt compression methods typically rely on auxiliary small language models (SLMs) to estimate context importance. However, such approaches introduce significant memory and computational overhead, which limits their deployment on resource-constrained edge devices. In this paper, we propose CORE, a two-stage sentence-level prompt compression method that eliminates the need for SLMs. In the first stage, CORE constructs an answer set via named entity recognition (NER) and a clue set via semantic matching. In the second stage, CORE refines the clue set using an orthogonal residual retrieval strategy and designs a spatial proximity-based metric to filter the answer set. The two sets are then combined to form the final compressed context. We implement CORE on an NVIDIA Jetson AGX Orin edge device and a Huawei Nova smartphone. Experimental results demonstrate that within a 2000-token budget, CORE improves accuracy by at least 30.19% compared to state-of-the-art baselines, while reducing memory usage by at least 50.47% and achieving at least 1.94 times speedup on the edge device. Moreover, compared to the state-of-the-art LLMLingua2 method, CORE achieves a substantial energy reduction of 95.74% on the smartphone, highlighting its practicality and generalizability for mobile deployments.

cs.CL↗

arXiv2026

ClearDepth: Enhanced Stereo Perception of Transparent Objects for Robotic Manipulation

Transparent object depth perception poses a challenge in everyday life and logistics, primarily due to the inability of standard 3D sensors to accurately capture depth on transparent or reflective surfaces. This limitation significantly affects depth map and point cloud-reliant applications, especially in robotic manipulation. We developed a vision transformer-based algorithm for stereo depth recovery of transparent objects. This approach is complemented by an innovative feature post-fusion module, which enhances the accuracy of depth recovery by structural features in images. To address the high costs associated with dataset collection for stereo camera-based perception of transparent objects, our method incorporates a parameter-aligned, domain-adaptive, and physically realistic Sim2Real simulation for efficient data generation, accelerated by AI algorithm. Our experimental results demonstrate the model's exceptional Sim2Real generalizability in real-world scenarios, enabling precise depth mapping of transparent objects to assist in robotic manipulation. Project details are available at https://sites.google.com/view/cleardepth/ .

cs.RO↗

arXiv2025

Improving LLM Reasoning via Dependency-Aware Query Decomposition and Logic-Parallel Content Expansion

The integration of Large Language Models (LLMs) into real-time Web applications, such as AI-powered search and conversational agents, presents a fundamental Web infrastructure challenge: reconciling the demand for high-quality, complex reasoning with the stringent low-latency and high-throughput requirements of interactive services. Current LLM reasoning, hindered by computationally inefficient sequential generation and rigid reasoning strategies, creates a critical bottleneck for the Web services. Existing approaches typically optimize the LLM reasoning for either efficiency or quality but struggle to achieve both, and thus fail to meet the dual requirements of modern Web platforms. To overcome these limitations, we propose Orion, a novel and efficient reasoning framework that enables dependency-aware query decomposition and logic-parallel content expansion. Concretely, Orion decomposes a single query reasoning process into two synergistic phases: (1) \textit{key point generation}, which distills logically structured key points through retrieval-augmented few-shot prompting, and (2) \textit{content parallel expansion}, which concurrently elaborates on these points based on a dependency graph to ensure logical consistency. Furthermore, Orion introduces a pipeline scheduling mechanism that exploits the complementary computational characteristics of the two phases (generation imposes pressure on GPU computing and expansion stresses on GPU memory) across multiple queries, enabling cross-query parallelism and dramatically improving reasoning performance (\ie, efficiency and quality). Experiments on diverse benchmarks show that Orion not only delivers up to 4.33x higher token generation speed and 3.42x lower answer latency over the baselines but also improves reasoning quality by up to 18.75% through explicitly modeling inter-point dependencies.

cs.AI↗

arXiv2025

SABlock: Semantic-Aware KV Cache Eviction with Adaptive Compression Block Size

The growing memory footprint of the Key-Value (KV) cache poses a severe scalability bottleneck for long-context Large Language Model (LLM) inference. While KV cache eviction has emerged as an effective solution by discarding less critical tokens, existing token-, block-, and sentence-level compression methods struggle to balance semantic coherence and memory efficiency. To this end, we introduce SABlock, a \underline{s}emantic-aware KV cache eviction framework with \underline{a}daptive \underline{block} sizes. Specifically, SABlock first performs semantic segmentation to align compression boundaries with linguistic structures, then applies segment-guided token scoring to refine token importance estimation. Finally, for each segment, a budget-driven search strategy adaptively determines the optimal block size that preserves semantic integrity while improving compression efficiency under a given cache budget. Extensive experiments on long-context benchmarks demonstrate that SABlock consistently outperforms state-of-the-art baselines under the same memory budgets. For instance, on Needle-in-a-Haystack (NIAH), SABlock achieves 99.9% retrieval accuracy with only 96 KV entries, nearly matching the performance of the full-cache baseline that retains up to 8K entries. Under a fixed cache budget of 1,024, SABlock further reduces peak memory usage by 46.28% and achieves up to 9.5x faster decoding on a 128K context length.

cs.CL↗

arXiv2025

BioDet: Boosting Industrial Object Detection with Image Preprocessing Strategies

Accurate 6D pose estimation is essential for robotic manipulation in industrial environments. Existing pipelines typically rely on off-the-shelf object detectors followed by cropping and pose refinement, but their performance degrades under challenging conditions such as clutter, poor lighting, and complex backgrounds, making detection the critical bottleneck. In this work, we introduce a standardized and plug-in pipeline for 2D detection of unseen objects in industrial settings. Based on current SOTA baselines, our approach reduces domain shift and background artifacts through low-light image enhancement and background removal guided by open-vocabulary detection with foundation models. This design suppresses the false positives prevalent in raw SAM outputs, yielding more reliable detections for downstream pose estimation. Extensive experiments on real-world industrial bin-picking benchmarks from BOP demonstrate that our method significantly boosts detection accuracy while incurring negligible inference overhead, showing the effectiveness and practicality of the proposed method.

cs.CV↗

arXiv2025

Accelerating Mixture-of-Expert Inference with Adaptive Expert Split Mechanism

Mixture-of-Experts (MoE) has emerged as a promising architecture for modern large language models (LLMs). However, massive parameters impose heavy GPU memory (i.e., VRAM) demands, hindering the widespread adoption of MoE LLMs. Offloading the expert parameters to CPU RAM offers an effective way to alleviate the VRAM requirements for MoE inference. Existing approaches typically cache a small subset of experts in VRAM and dynamically prefetch experts from RAM during inference, leading to significant degradation in inference speed due to the poor cache hit rate and substantial expert loading latency. In this work, we propose MoEpic, an efficient MoE inference system with a novel expert split mechanism. Specifically, each expert is vertically divided into two segments: top and bottom. MoEpic caches the top segment of hot experts, so that more experts will be stored under the limited VRAM budget, thereby improving the cache hit rate. During each layer's inference, MoEpic predicts and prefetches the activated experts for the next layer. Since the top segments of cached experts are exempt from fetching, the loading time is reduced, which allows efficient transfer-computation overlap. Nevertheless, the performance of MoEpic critically depends on the cache configuration (i.e., each layer's VRAM budget and expert split ratio). To this end, we propose a divide-and-conquer algorithm based on fixed-point iteration for adaptive cache configuration. Extensive experiments on popular MoE LLMs demonstrate that MoEpic can save about half of the GPU cost, while lowering the inference latency by about 37.51%-65.73% compared to the baselines.

cs.LG↗

arXiv2025

Towards Communication-Efficient Decentralized Federated Graph Learning over Non-IID Data

Decentralized Federated Graph Learning (DFGL) overcomes potential bottlenecks of the parameter server in FGL by establishing a peer-to-peer (P2P) communication network among workers. However, while extensive cross-worker communication of graph node embeddings is crucial for DFGL training, it introduces substantial communication costs. Most existing works typically construct sparse network topologies or utilize graph neighbor sampling methods to alleviate the communication overhead in DFGL. Intuitively, integrating these methods may offer promise for doubly improving communication efficiency in DFGL. However, our preliminary experiments indicate that directly combining these methods leads to significant training performance degradation if they are jointly optimized. To address this issue, we propose Duplex, a unified framework that jointly optimizes network topology and graph sampling by accounting for their coupled relationship, thereby significantly reducing communication cost while enhancing training performance in DFGL. To overcome practical DFGL challenges, eg, statistical heterogeneity and dynamic network environments, Duplex introduces a learning-driven algorithm to adaptively determine optimal network topologies and graph sampling ratios for workers. Experimental results demonstrate that Duplex reduces completion time by 20.1%--48.8% and communication costs by 16.7%--37.6% to achieve target accuracy, while improving accuracy by 3.3%--7.9% under identical resource budgets compared to baselines.

cs.DC↗

arXiv2025

Adaptive KV-Cache Compression without Manually Setting Budget

Large language models (LLMs) inference relies heavily on KV-caches to accelerate autoregressive decoding, but the resulting memory footprint grows rapidly with sequence length, posing significant efficiency challenges. Current KV-cache compression methods suffer from a Procrustes' bed problem: they force diverse workloads into fixed compression ratios, leading to suboptimal resource allocation and inference performance. To this end, we present GVote, an adaptive KV-cache compression scheme that eliminates manual budget specification while achieving superior accuracy-efficiency trade-offs. GVote operates on the principle that the important keys are the aggregation of keys required by future queries. The method predicts future query attention demands by Monte-Carlo style sampling potential queries and aggregating selected keys to determine the optimal cache budget without manual specification. Experimental evaluation demonstrates GVote's effectiveness across multiple benchmarks, including GSM8K, RULER and Longbench. Compared to baselines, GVote exhibits 2$\times$ memory reduction while the accuracy maintains higher or comparable.

cs.DB↗

arXiv2025

DySTop

Federated Learning (FL) has emerged as a potential distributed learning paradigm that enables model training on edge devices (i.e., workers) while preserving data privacy. However, its reliance on a centralized server leads to limited scalability. Decentralized federated learning (DFL) eliminates the dependency on a centralized server by enabling peer-to-peer model exchange. Existing DFL mechanisms mainly employ synchronous communication, which may result in training inefficiencies under heterogeneous and dynamic edge environments. Although a few recent asynchronous DFL (ADFL) mechanisms have been proposed to address these issues, they typically yield stale model aggregation and frequent model transmission, leading to degraded training performance on non-IID data and high communication overhead. To overcome these issues, we present DySTop, an innovative mechanism that jointly optimizes dynamic staleness control and topology construction in ADFL. In each round, multiple workers are activated, and a subset of their neighbors is selected to transmit models for aggregation, followed by local training. We provide a rigorous convergence analysis for DySTop, theoretically revealing the quantitative relationships between the convergence bound and key factors such as maximum staleness, activating frequency, and data distribution among workers. From the insights of the analysis, we propose a worker activation algorithm (WAA) for staleness control and a phase-aware topology construction algorithm (PTCA) to reduce communication overhead and handle data non-IID. Extensive evaluations through both large-scale simulations and real-world testbed experiments demonstrate that our DySTop reduces completion time by 51.8% and the communication resource consumption by 57.1% compared to state-of-the-art solutions, while maintaining the same model accuracy.

cs.DC↗

arXiv2025

Air-FedGA: A Grouping Asynchronous Federated Learning Mechanism Exploiting Over-the-air Computation

Federated learning (FL) is a new paradigm to train AI models over distributed edge devices (i.e., workers) using their local data, while confronting various challenges including communication resource constraints, edge heterogeneity and data Non-IID. Over-the-air computation (AirComp) is a promising technique to achieve efficient utilization of communication resource for model aggregation by leveraging the superposition property of a wireless multiple access channel (MAC). However, AirComp requires strict synchronization among edge devices, which is hard to achieve in heterogeneous scenarios. In this paper, we propose an AirComp-based grouping asynchronous federated learning mechanism (Air-FedGA), which combines the advantages of AirComp and asynchronous FL to address the communication and heterogeneity challenges. Specifically, Air-FedGA organizes workers into groups and performs over-the-air aggregation within each group, while groups asynchronously communicate with the parameter server to update the global model. In this way, Air-FedGA accelerates the FL model training by over-the-air aggregation, while relaxing the synchronization requirement of this aggregation technology. We theoretically prove the convergence of Air-FedGA. We formulate a training time minimization problem for Air-FedGA and propose the power control and worker grouping algorithm to solve it, which jointly optimizes the power scaling factors at edge devices, the denoising factors at the parameter server, as well as the worker grouping strategy. We conduct experiments on classical models and datasets, and the results demonstrate that our proposed mechanism and algorithm can speed up FL model training by 29.9%-71.6% compared with the state-of-the-art solutions.

cs.DC↗

arXiv2025

GCE-Pose: Global Context Enhancement for Category-level Object Pose Estimation

A key challenge in model-free category-level pose estimation is the extraction of contextual object features that generalize across varying instances within a specific category. Recent approaches leverage foundational features to capture semantic and geometry cues from data. However, these approaches fail under partial visibility. We overcome this with a first-complete-then-aggregate strategy for feature extraction utilizing class priors. In this paper, we present GCE-Pose, a method that enhances pose estimation for novel instances by integrating category-level global context prior. GCE-Pose performs semantic shape reconstruction with a proposed Semantic Shape Reconstruction (SSR) module. Given an unseen partial RGB-D object instance, our SSR module reconstructs the instance's global geometry and semantics by deforming category-specific 3D semantic prototypes through a learned deep Linear Shape Model. We further introduce a Global Context Enhanced (GCE) feature fusion module that effectively fuses features from partial RGB-D observations and the reconstructed global context. Extensive experiments validate the impact of our global context prior and the effectiveness of the GCE fusion module, demonstrating that GCE-Pose significantly outperforms existing methods on challenging real-world datasets HouseCat6D and NOCS-REAL275. Our project page is available at https://colin-de.github.io/GCE-Pose/.

cs.CV↗

arXiv2025

Mitigating Catastrophic Forgetting with Adaptive Transformer Block Expansion in Federated Fine-Tuning

Federated fine-tuning (FedFT) of large language models (LLMs) has emerged as a promising solution for adapting models to distributed data environments while ensuring data privacy. Existing FedFT methods predominantly utilize parameter-efficient fine-tuning (PEFT) techniques to reduce communication and computation overhead. However, they often fail to adequately address the catastrophic forgetting, a critical challenge arising from continual adaptation in distributed environments. The traditional centralized fine-tuning methods, which are not designed for the heterogeneous and privacy-constrained nature of federated environments, struggle to mitigate this issue effectively. Moreover, the challenge is further exacerbated by significant variation in data distributions and device capabilities across clients, which leads to intensified forgetting and degraded model generalization. To tackle these issues, we propose FedBE, a novel FedFT framework that integrates an adaptive transformer block expansion mechanism with a dynamic trainable-block allocation strategy. Specifically, FedBE expands trainable blocks within the model architecture, structurally separating newly learned task-specific knowledge from the original pre-trained representations. Additionally, FedBE dynamically assigns these trainable blocks to clients based on their data distributions and computational capabilities. This enables the framework to better accommodate heterogeneous federated environments and enhances the generalization ability of the model.Extensive experiments show that compared with existing federated fine-tuning methods, FedBE achieves 12-74% higher accuracy retention on general tasks after fine-tuning and a model convergence acceleration ratio of 1.9-3.1x without degrading the accuracy of downstream tasks.

cs.LG↗

arXiv2025

FedQuad: Adaptive Layer-wise LoRA Deployment and Activation Quantization for Federated Fine-Tuning

Federated fine-tuning (FedFT) provides an effective paradigm for fine-tuning large language models (LLMs) in privacy-sensitive scenarios. However, practical deployment remains challenging due to the limited resources on end devices. Existing methods typically utilize parameter-efficient fine-tuning (PEFT) techniques, such as Low-Rank Adaptation (LoRA), to substantially reduce communication overhead. Nevertheless, significant memory usage for activation storage and computational demands from full backpropagation remain major barriers to efficient deployment on resource-constrained end devices. Moreover, substantial resource heterogeneity across devices results in severe synchronization bottlenecks, diminishing the overall fine-tuning efficiency. To address these issues, we propose FedQuad, a novel LoRA-based FedFT framework that adaptively adjusts the LoRA depth (the number of consecutive tunable LoRA layers from the output) according to device computational capabilities, while employing activation quantization to reduce memory overhead, thereby enabling efficient deployment on resource-constrained devices. Specifically, FedQuad first identifies the feasible and efficient combinations of LoRA depth and the number of activation quantization layers based on device-specific resource constraints. Subsequently, FedQuad employs a greedy strategy to select the optimal configurations for each device, effectively accommodating system heterogeneity. Extensive experiments demonstrate that FedQuad achieves a 1.4-5.3x convergence acceleration compared to state-of-the-art baselines when reaching target accuracy, highlighting its efficiency and deployability in resource-constrained and heterogeneous end-device environments.

cs.DC↗

arXiv2025

Collaborative Speculative Inference for Efficient LLM Inference Serving

Speculative inference is a promising paradigm employing small speculative models (SSMs) as drafters to generate draft tokens, which are subsequently verified in parallel by the target large language model (LLM). This approach enhances the efficiency of inference serving by reducing LLM inference latency and costs while preserving generation quality. However, existing speculative methods face critical challenges, including inefficient resource utilization and limited draft acceptance, which constrain their scalability and overall effectiveness. To overcome these obstacles, we present CoSine, a novel speculative inference system that decouples sequential speculative decoding from parallel verification, enabling efficient collaboration among multiple nodes. Specifically, CoSine routes inference requests to specialized drafters based on their expertise and incorporates a confidence-based token fusion mechanism to synthesize outputs from cooperating drafters, ensuring high-quality draft generation. Additionally, CoSine dynamically orchestrates the execution of speculative decoding and verification in a pipelined manner, employing batch scheduling to selectively group requests and adaptive speculation control to minimize idle periods. By optimizing parallel workflows through heterogeneous node collaboration, CoSine balances draft generation and verification throughput in real-time, thereby maximizing resource utilization. Experimental results demonstrate that CoSine achieves superior performance compared to state-of-the-art speculative approaches. Notably, with equivalent resource costs, CoSine achieves up to a 23.2% decrease in latency and a 32.5% increase in throughput compared to baseline methods.

cs.DC↗

arXiv2025

Cross-region Model Training with Communication-Computation Overlapping and Delay Compensation

Training large language models (LLMs) requires massive computational resources, often necessitating the aggregation of geographically distributed data centers (\ie, cross-region training). However, the high communication latency in wide-area networks severely degrades the efficiency of traditional distributed training. While methods like DiLoCo reduce communication frequency, they suffer from blocking synchronization. Streaming DiLoCo alleviates this issue via communication-computation overlapping but introduces update staleness and model inconsistency due to delayed global updates and partial synchronization. These factors impair convergence, especially when aggressive overlap is needed to mask high latency. We propose CoCoDC, a novel distributed training framework with communication-computation overlapping and delay compensation, to explicitly tackle these challenges. Within the CoCoDC framework, we specifically develop a novel Delay Compensation strategy based on Taylor expansion to effectively mitigate the staleness and an Adaptive Transmission strategy that dynamically schedules model fragment synchronization to optimize bandwidth usage and accelerate convergence. Extensive experiments highlight the superior performance of CoCoDC over both DiLoCo and Streaming DiLoCo regarding final accuracy and training speed. Specifically, CoCoDC reduces the training steps needed to reach a comparable perplexity by up to 21.0% compared to Streaming DiLoCo. Our work provides an effective solution for scalable and efficient cross-region LLM training.

cs.DC↗

arXiv2025

PRISM: Probabilistic Representation for Integrated Shape Modeling and Generation

Despite the advancements in 3D full-shape generation, accurately modeling complex geometries and semantics of shape parts remains a significant challenge, particularly for shapes with varying numbers of parts. Current methods struggle to effectively integrate the contextual and structural information of 3D shapes into their generative processes. We address these limitations with PRISM, a novel compositional approach for 3D shape generation that integrates categorical diffusion models with Statistical Shape Models (SSM) and Gaussian Mixture Models (GMM). Our method employs compositional SSMs to capture part-level geometric variations and uses GMM to represent part semantics in a continuous space. This integration enables both high fidelity and diversity in generated shapes while preserving structural coherence. Through extensive experiments on shape generation and manipulation tasks, we demonstrate that our approach significantly outperforms previous methods in both quality and controllability of part-level operations. Our code will be made publicly available.

cs.CV↗