arXiv subjects

Li Jiang

Publications and source records attributed to Li Jiang.

At least 55 records · Page 3Linked to original sources

arXiv2026

FlashVID: Efficient Video Large Language Models via Training-free Tree-based Spatiotemporal Token Merging

Although Video Large Language Models (VLLMs) have shown remarkable capabilities in video understanding, they are required to process high volumes of visual tokens, causing significant computational inefficiency. Existing VLLMs acceleration frameworks usually compress spatial and temporal redundancy independently, which overlooks the spatiotemporal relationships, thereby leading to suboptimal spatiotemporal compression. The highly correlated visual features are likely to change in spatial position, scale, orientation, and other attributes over time due to the dynamic nature of video. Building on this insight, we introduce FlashVID, a training-free inference acceleration framework for VLLMs. Specifically, FlashVID utilizes Attention and Diversity-based Token Selection (ADTS) to select the most representative tokens for basic video representation, then applies Tree-based Spatiotemporal Token Merging (TSTM) for fine-grained spatiotemporal redundancy elimination. Extensive experiments conducted on three representative VLLMs across five video understanding benchmarks demonstrate the effectiveness and generalization of our method. Notably, by retaining only 10% of visual tokens, FlashVID preserves 99.1% of the performance of LLaVA-OneVision. Consequently, FlashVID can serve as a training-free and plug-and-play module for extending long video frames, which enables a 10x increase in video frame input to Qwen2.5-VL, resulting in a relative improvement of 8.6% within the same computational budget. Code is available at https://github.com/Fanziyang-v/FlashVID.

cs.CV↗

arXiv2026

Proteus-ID: ID-Consistent and Motion-Coherent Video Customization

Video identity customization seeks to synthesize realistic, temporally coherent videos of a specific subject, given a single reference image and a text prompt. This task presents two core challenges: (1) maintaining identity consistency while aligning with the described appearance and actions, and (2) generating natural, fluid motion without unrealistic stiffness. To address these challenges, we introduce Proteus-ID, a novel diffusion-based framework for identity-consistent and motion-coherent video customization. First, we propose a Multimodal Identity Fusion (MIF) module that unifies visual and textual cues into a joint identity representation using a Q-Former, providing coherent guidance to the diffusion model and eliminating modality imbalance. Second, we present a Time-Aware Identity Injection (TAII) mechanism that dynamically modulates identity conditioning across denoising steps, improving fine-detail reconstruction. Third, we propose Adaptive Motion Learning (AML), a self-supervised strategy that reweights the training loss based on optical-flow-derived motion heatmaps, enhancing motion realism without requiring additional inputs. To support this task, we construct Proteus-Bench, a high-quality dataset comprising 200K curated clips for training and 150 individuals from diverse professions and ethnicities for evaluation. Extensive experiments demonstrate that Proteus-ID outperforms prior methods in identity preservation, text alignment, and motion quality, establishing a new benchmark for video identity customization. Codes and data are publicly available at https://grenoble-zhang.github.io/Proteus-ID/.

cs.CV↗

arXiv2026

HyperOffload: Graph-Driven Hierarchical Memory Management for Large Language Models on SuperNode Architectures

The rapid evolution of Large Language Models (LLMs) towards long-context reasoning and sparse architectures has pushed memory requirements far beyond the capacity of individual device HBM. While emerging supernode architectures offer terabyte-scale shared memory pools via high-bandwidth interconnects, existing software stacks fail to exploit this hardware effectively. Current runtime-based offloading and swapping techniques operate with a local view, leading to reactive scheduling and exposed communication latency that stall the computation pipeline. In this paper, we propose the SuperNode Memory Management Framework (\textbf{HyperOffload}). It employs a compiler-assisted approach that leverages graph-driven memory management to treat remote memory access as explicit operations in the computation graph, specifically designed for hierarchical SuperNode architectures. Unlike reactive runtime systems, SuperNode represents data movement using cache operators within the compiler's Intermediate Representation (IR). This design enables a global, compile-time analysis of tensor lifetimes and execution dependencies. Leveraging this visibility, we develop a global execution-order refinement algorithm that statically schedules data transfers to hide remote memory latency behind compute-intensive regions. We implement SuperNode within the production deep learning framework MindSpore, adding a remote memory backend and specialized compiler passes. Evaluation on representative LLM workloads shows that SuperNode reduces peak device memory usage by up to 26\% for inference while maintaining end-to-end performance. Our work demonstrates that integrating memory-augmented hardware into the compiler's optimization framework is essential for scaling next-generation AI workloads.

cs.DC↗

arXiv2026

LIVE: Long-horizon Interactive Video World Modeling

Autoregressive video world models predict future visual observations conditioned on actions. While effective over short horizons, these models often struggle with long-horizon generation, as small prediction errors accumulate over time. Prior methods alleviate this by introducing pre-trained teacher models and sequence-level distribution matching, which incur additional computational cost and fail to prevent error propagation beyond the training horizon. In this work, we propose LIVE, a Long-horizon Interactive Video world modEl that enforces bounded error accumulation via a novel cycle-consistency objective, thereby eliminating the need for teacher-based distillation. Specifically, LIVE first performs a forward rollout from ground-truth frames and then applies a reverse generation process to reconstruct the initial state. The diffusion loss is subsequently computed on the reconstructed terminal state, providing an explicit constraint on long-horizon error propagation. Moreover, we provide an unified view that encompasses different approaches and introduce progressive training curriculum to stabilize training. Experiments demonstrate that LIVE achieves state-of-the-art performance on long-horizon benchmarks, generating stable, high-quality videos far beyond training rollout lengths.

cs.CV↗

arXiv2025

Are Bus-Mounted Edge Servers Feasible?

Placement of edge servers is the prerequisite of provisioning edge computing services for Internet of Vehicles (IoV). Fixed-site edge servers at Road Side Units (RSUs) or base stations are able to offer basic service coverage for end users, i.e., vehicles on road. However, the server locations and capacity are fixed after deployment, rendering their inefficiency in handling spationtemporal user dynamics. Mobile servers such as buses, on the other hand, have the potential of adding computation elasticity to such system. To this end, this paper studies the feasibility of bus-mounted edge servers based on real traces. First, we investigate the coverage of the buses and base stations using the Shanghai bus/taxi/Telecom datasets, which shows a great potential of bus-based edge servers as they cover a great portion of geographic area and demand points. Next, we build a mathematical model and design a simple greedy heuristic algorithm to select a limited number of buses that maximizes the coverage of demand points, i.e., with a limited purchase budget. We perform trace-driven simulations to verify the performance of the proposed bus selection algorithm. The results show that our approach effectively handles the dynamic user demand under realistic constraints such as server capacity and purchase quantity. Thus, we claim: bus-mounted edge servers for vehicular networks in urban areas are feasible, beneficial, and valuable.

cs.DC↗

arXiv2025

UniSplat: Unified Spatio-Temporal Fusion via 3D Latent Scaffolds for Dynamic Driving Scene Reconstruction

Feed-forward 3D reconstruction for autonomous driving has advanced rapidly, yet existing methods struggle with the joint challenges of sparse, non-overlapping camera views and complex scene dynamics. We present UniSplat, a general feed-forward framework that learns robust dynamic scene reconstruction through unified latent spatio-temporal fusion. UniSplat constructs a 3D latent scaffold, a structured representation that captures geometric and semantic scene context by leveraging pretrained foundation models. To effectively integrate information across spatial views and temporal frames, we introduce an efficient fusion mechanism that operates directly within the 3D scaffold, enabling consistent spatio-temporal alignment. To ensure complete and detailed reconstructions, we design a dual-branch decoder that generates dynamic-aware Gaussians from the fused scaffold by combining point-anchored refinement with voxel-based generation, and maintain a persistent memory of static Gaussians to enable streaming scene completion beyond current camera coverage. Extensive experiments on real-world datasets demonstrate that UniSplat achieves state-of-the-art performance in novel view synthesis, while providing robust and high-quality renderings even for viewpoints outside the original camera coverage.

cs.CV↗

arXiv2025

LaMoS: Enabling Efficient Large Number Modular Multiplication through SRAM-based CiM Acceleration

Barrett's algorithm is one of the most widely used methods for performing modular multiplication, a critical nonlinear operation in modern privacy computing techniques such as homomorphic encryption (HE) and zero-knowledge proofs (ZKP). Since modular multiplication dominates the processing time in these applications, computational complexity and memory limitations significantly impact performance. Computing-in-Memory (CiM) is a promising approach to tackle this problem. However, existing schemes currently suffer from two main problems: 1) Most works focus on low bit-width modular multiplication, which is inadequate for mainstream cryptographic algorithms such as elliptic curve cryptography (ECC) and the RSA algorithm, both of which require high bit-width operations; 2) Recent efforts targeting large number modular multiplication rely on inefficient in-memory logic operations, resulting in high scaling costs for larger bit-widths and increased latency. To address these issues, we propose LaMoS, an efficient SRAM-based CiM design for large-number modular multiplication, offering high scalability and area efficiency. First, we analyze the Barrett's modular multiplication method and map the workload onto SRAM CiM macros for high bit-width cases. Additionally, we develop an efficient CiM architecture and dataflow to optimize large-number modular multiplication. Finally, we refine the mapping scheme for better scalability in high bit-width scenarios using workload grouping. Experimental results show that LaMoS achieves a $7.02\times$ speedup and reduces high bit-width scaling costs compared to existing SRAM-based CiM designs.

cs.CR↗

arXiv2025

Hummer: Towards Limited Competitive Preference Dataset

Preference datasets are essential for incorporating human preferences into pre-trained language models, playing a key role in the success of Reinforcement Learning from Human Feedback. However, these datasets often demonstrate conflicting alignment objectives, leading to increased vulnerability to jailbreak attacks and challenges in adapting downstream tasks to prioritize specific alignment objectives without negatively impacting others. In this work, we introduce a novel statistical metric, Alignment Dimension Conflict, to quantify the degree of conflict within preference datasets. We then present \texttt{Hummer} and its fine-grained variant, \texttt{Hummer-F}, as innovative pairwise preference datasets with reduced-conflict alignment objectives. \texttt{Hummer} is built based on UltraFeedback and is enhanced by AI feedback from GPT-4, marking as the first preference dataset aimed at reducing the competition between alignment objectives. Furthermore, we develop reward models, HummerRM and HummerRM-F, which employ a hybrid sampling approach to balance diverse alignment objectives effectively. This sampling method positions HummerRM as an ideal model for domain-specific further fine-tuning and reducing vulnerabilities to attacks.

cs.AI↗

arXiv2025

FlexQuant: A Flexible and Efficient Dynamic Precision Switching Framework for LLM Quantization

The rapid advancement of large language models (LLMs) has exacerbated the memory bottleneck due to the widening gap between model parameter scaling and hardware capabilities. While post-training quantization techniques effectively reduce memory overhead, existing methods predominantly rely on static quantization strategies, which struggle to adapt to dynamic workloads. To address this, we propose FlexQuant, a dynamic precision-switching framework that optimizes the trade-off between inference speed and accuracy. Leveraging model perplexity entropy and Kullback-Leibler divergence, FlexQuant enables fine-grained, layer-wise mixed-precision quantization and dynamically adjusts bit-widths during each token generation. FlexQuant provides a comprehensive analysis of quantization strategies, introduces a precision requirement model for optimal switching, and implements efficient fine-grained precision management. Evaluations demonstrate that FlexQuant achieves a 1.3x end-to-end speedup across diverse language tasks with negligible accuracy loss introduced. This framework offers a flexible and adaptive solution for efficient LLM deployment. Code is released at https://github.com/ZongwuWang/FlexQuant.git.

cs.LG↗

arXiv2025

Memory Forcing: Spatio-Temporal Memory for Consistent Scene Generation on Minecraft

Autoregressive video diffusion models have proved effective for world modeling and interactive scene generation, with Minecraft gameplay as a representative application. To faithfully simulate play, a model must generate natural content while exploring new scenes and preserve spatial consistency when revisiting explored areas. Under limited computation budgets, it must compress and exploit historical cues within a finite context window, which exposes a trade-off: Temporal-only memory lacks long-term spatial consistency, whereas adding spatial memory strengthens consistency but may degrade new scene generation quality when the model over-relies on insufficient spatial context. We present Memory Forcing, a learning framework that pairs training protocols with a geometry-indexed spatial memory. Hybrid Training exposes distinct gameplay regimes, guiding the model to rely on temporal memory during exploration and incorporate spatial memory for revisits. Chained Forward Training extends autoregressive training with model rollouts, where chained predictions create larger pose variations and encourage reliance on spatial memory for maintaining consistency. Point-to-Frame Retrieval efficiently retrieves history by mapping currently visible points to their source frames, while Incremental 3D Reconstruction maintains and updates an explicit 3D cache. Extensive experiments demonstrate that Memory Forcing achieves superior long-term spatial consistency and generative quality across diverse environments, while maintaining computational efficiency for extended sequences.

cs.CV↗

arXiv2025

AGS: Accelerating 3D Gaussian Splatting SLAM via CODEC-Assisted Frame Covisibility Detection

Simultaneous Localization and Mapping (SLAM) is a critical task that enables autonomous vehicles to construct maps and localize themselves in unknown environments. Recent breakthroughs combine SLAM with 3D Gaussian Splatting (3DGS) to achieve exceptional reconstruction fidelity. However, existing 3DGS-SLAM systems provide insufficient throughput due to the need for multiple training iterations per frame and the vast number of Gaussians. In this paper, we propose AGS, an algorithm-hardware co-design framework to boost the efficiency of 3DGS-SLAM based on the intuition that SLAM systems process frames in a streaming manner, where adjacent frames exhibit high similarity that can be utilized for acceleration. On the software level: 1) We propose a coarse-then-fine-grained pose tracking method with respect to the robot's movement. 2) We avoid redundant computations of Gaussians by sharing their contribution information across frames. On the hardware level, we propose a frame covisibility detection engine to extract intermediate data from the video CODEC. We also implement a pose tracking engine and a mapping engine with workload schedulers to efficiently deploy the AGS algorithm. Our evaluation shows that AGS achieves up to $17.12\times$, $6.71\times$, and $5.41\times$ speedups against the mobile and high-end GPUs, and a state-of-the-art 3DGS accelerator, GSCore.

cs.AR↗

arXiv2025

ASDR: Exploiting Adaptive Sampling and Data Reuse for CIM-based Instant Neural Rendering

Neural Radiance Fields (NeRF) offer significant promise for generating photorealistic images and videos. However, existing mainstream neural rendering models often fall short in meeting the demands for immediacy and power efficiency in practical applications. Specifically, these models frequently exhibit irregular access patterns and substantial computational overhead, leading to undesirable inference latency and high power consumption. Computing-in-memory (CIM), an emerging computational paradigm, has the potential to address these access bottlenecks and reduce the power consumption associated with model execution. To bridge the gap between model performance and real-world scene requirements, we propose an algorithm-architecture co-design approach, abbreviated as ASDR, a CIM-based accelerator supporting efficient neural rendering. At the algorithmic level, we propose two rendering optimization schemes: (1) Dynamic sampling by online sensing of the rendering difficulty of different pixels, thus reducing access memory and computational overhead. (2) Reducing MLP overhead by decoupling and approximating the volume rendering of color and density. At the architecture level, we design an efficient ReRAM-based CIM architecture with efficient data mapping and reuse microarchitecture. Experiments demonstrate that our design can achieve up to $9.55\times$ and $69.75\times$ speedup over state-of-the-art NeRF accelerators and Xavier NX GPU in graphics rendering tasks with only $0.1$ PSNR loss.

cs.AR↗

arXiv2025

PUSHtap: PIM-based In-Memory HTAP with Unified Data Storage Format

Hybrid transaction/analytical processing (HTAP) is an emerging database paradigm that supports both online transaction processing (OLTP) and online analytical processing (OLAP) workloads. Computing-intensive OLTP operations, involving row-wise data manipulation, are suitable for row-store format. In contrast, memory-intensive OLAP operations, which are column-centric, benefit from column-store format. This \emph{data-format dilemma} prevents HTAP systems from concurrently achieving three design goals: performance isolation, data freshness, and workload-specific optimization. Another background technology is Processing-in-Memory (PIM), which integrates computing units (PIM units) inside DRAM memory devices to accelerate memory-intensive workloads, including OLAP. Our key insight is to combine the interleaved CPU access and localized PIM unit access to provide two-dimensional access to address the data format contradictions inherent in HTAP. First, we propose a unified data storage format with novel data alignment and placement techniques to optimize the effective bandwidth of CPUs and PIM units and exploit the PIM's parallelism. Second, we implement the multi-version concurrency control (MVCC) essential for single-instance HTAP. Third, we extend the commercial PIM architecture to support the OLAP operations and concurrent access from PIM and CPU. Experiments show that PUSHtap can achieve 3.4\texttimes{}/4.4\texttimes{} OLAP/OLTP throughput improvement compared to multi-instance PIM-based design.

cs.DC↗

arXiv2025

Edit360: 2D Image Edits to 3D Assets from Any Angle

Recent advances in diffusion models have significantly improved image generation and editing, but extending these capabilities to 3D assets remains challenging, especially for fine-grained edits that require multi-view consistency. Existing methods typically restrict editing to predetermined viewing angles, severely limiting their flexibility and practical applications. We introduce Edit360, a tuning-free framework that extends 2D modifications to multi-view consistent 3D editing. Built upon video diffusion models, Edit360 enables user-specific editing from arbitrary viewpoints while ensuring structural coherence across all views. The framework selects anchor views for 2D modifications and propagates edits across the entire 360-degree range. To achieve this, Edit360 introduces a novel Anchor-View Editing Propagation mechanism, which effectively aligns and merges multi-view information within the latent and attention spaces of diffusion models. The resulting edited multi-view sequences facilitate the reconstruction of high-quality 3D assets, enabling customizable 3D content creation.

cs.GR↗

arXiv2025

Enhancing Spatial Reasoning in Multimodal Large Language Models through Reasoning-based Segmentation

Recent advances in point cloud perception have demonstrated remarkable progress in scene understanding through vision-language alignment leveraging large language models (LLMs). However, existing methods may still encounter challenges in handling complex instructions that require accurate spatial reasoning, even if the 3D point cloud data provides detailed spatial cues such as size and position for identifying the targets. To tackle this issue, we propose Relevant Reasoning Segmentation (R$^2$S), a reasoning-based segmentation framework. The framework emulates human cognitive processes by decomposing spatial reasoning into two sequential stages: first identifying relevant elements, then processing instructions guided by their associated visual priors. Furthermore, acknowledging the inadequacy of existing datasets in complex reasoning tasks, we introduce 3D ReasonSeg, a reasoning-based segmentation dataset comprising 25,185 training samples and 3,966 validation samples with precise annotations. Both quantitative and qualitative experiments demonstrate that the R$^2$S and 3D ReasonSeg effectively endow 3D point cloud perception with stronger spatial reasoning capabilities, and we hope that they can serve as a new baseline and benchmark for future work.

cs.CV↗

arXiv2025

ALLMod: Exploring $\underline{\mathbf{A}}$rea-Efficiency of $\underline{\mathbf{L}}$UT-based $\underline{\mathbf{L}}$arge Number $\underline{\mathbf{Mod}}$ular Reduction via Hybrid Workloads

Modular arithmetic, particularly modular reduction, is widely used in cryptographic applications such as homomorphic encryption (HE) and zero-knowledge proofs (ZKP). High-bit-width operations are crucial for enhancing security; however, they are computationally intensive due to the large number of modular operations required. The lookup-table-based (LUT-based) approach, a ``space-for-time'' technique, reduces computational load by segmenting the input number into smaller bit groups, pre-computing modular reduction results for each segment, and storing these results in LUTs. While effective, this method incurs significant hardware overhead due to extensive LUT usage. In this paper, we introduce ALLMod, a novel approach that improves the area efficiency of LUT-based large-number modular reduction by employing hybrid workloads. Inspired by the iterative method, ALLMod splits the bit groups into two distinct workloads, achieving lower area costs without compromising throughput. We first develop a template to facilitate workload splitting and ensure balanced distribution. Then, we conduct design space exploration to evaluate the optimal timing for fusing workload results, enabling us to identify the most efficient design under specific constraints. Extensive evaluations show that ALLMod achieves up to $1.65\times$ and $3\times$ improvements in area efficiency over conventional LUT-based methods for bit-widths of $128$ and $8,192$, respectively.

cs.CR↗

arXiv2025

DriveX: Omni Scene Modeling for Learning Generalizable World Knowledge in Autonomous Driving

Data-driven learning has advanced autonomous driving, yet task-specific models struggle with out-of-distribution scenarios due to their narrow optimization objectives and reliance on costly annotated data. We present DriveX, a self-supervised world model that learns generalizable scene dynamics and holistic representations (geometric, semantic, and motion) from large-scale driving videos. DriveX introduces Omni Scene Modeling (OSM), a module that unifies multimodal supervision-3D point cloud forecasting, 2D semantic representation, and image generation-to capture comprehensive scene evolution. To simplify learning complex dynamics, we propose a decoupled latent world modeling strategy that separates world representation learning from future state decoding, augmented by dynamic-aware ray sampling to enhance motion modeling. For downstream adaptation, we design Future Spatial Attention (FSA), a unified paradigm that dynamically aggregates spatiotemporal features from DriveX's predictions to enhance task-specific inference. Extensive experiments demonstrate DriveX's effectiveness: it achieves significant improvements in 3D future point cloud prediction over prior work, while attaining state-of-the-art results on diverse tasks including occupancy prediction, flow estimation, and end-to-end driving. These results validate DriveX's capability as a general-purpose world model, paving the way for robust and unified autonomous driving frameworks.

cs.CV↗

arXiv2025

DASH: Input-Aware Dynamic Layer Skipping for Efficient LLM Inference with Markov Decision Policies

Large language models (LLMs) have achieved remarkable performance across a wide range of NLP tasks. However, their substantial inference cost poses a major barrier to real-world deployment, especially in latency-sensitive scenarios. To address this challenge, we propose \textbf{DASH}, an adaptive layer-skipping framework that dynamically selects computation paths conditioned on input characteristics. We model the skipping process as a Markov Decision Process (MDP), enabling fine-grained token-level decisions based on intermediate representations. To mitigate potential performance degradation caused by skipping, we introduce a lightweight compensation mechanism that injects differential rewards into the decision process. Furthermore, we design an asynchronous execution strategy that overlaps layer computation with policy evaluation to minimize runtime overhead. Experiments on multiple LLM architectures and NLP benchmarks show that our method achieves significant inference acceleration while maintaining competitive task performance, outperforming existing methods.

cs.CL↗