arXiv subjects

Chenghao Li

Publications and source records attributed to Chenghao Li.

At least 37 records · Page 2Linked to original sources

arXiv2025

Syzygy of Thoughts: Improving LLM CoT with the Minimal Free Resolution

Chain-of-Thought (CoT) prompting enhances the reasoning of large language models (LLMs) by decomposing problems into sequential steps, mimicking human logic and reducing errors. However, complex tasks with vast solution spaces and vague constraints often exceed the capacity of a single reasoning chain. Inspired by Minimal Free Resolution (MFR) in commutative algebra and algebraic geometry, we propose Syzygy of Thoughts (SoT)-a novel framework that extends CoT by introducing auxiliary, interrelated reasoning paths. SoT captures deeper logical dependencies, enabling more robust and structured problem-solving. MFR decomposes a module into a sequence of free modules with minimal rank, providing a structured analytical approach to complex systems. This method introduces the concepts of "Module", "Betti numbers","Freeness", "Mapping", "Exactness" and "Minimality", enabling the systematic decomposition of the original complex problem into logically complete minimal subproblems while preserving key problem features and reducing reasoning length. We tested SoT across diverse datasets (e.g., GSM8K, MATH) and models (e.g., GPT-4o-mini, Qwen2.5), achieving inference accuracy that matches or surpasses mainstream CoTs standards. Additionally, by aligning the sampling process with algebraic constraints, our approach enhances the scalability of inference time in LLMs, ensuring both transparent reasoning and high performance. Our code will be publicly available at https://github.com/dlMARiA/Syzygy-of-thoughts.

cs.CL↗

arXiv2025

MCI: Multi-Channel Imager on the Chinese Space Station Survey Telescope

The Multi-Channel Imager (MCI) is a powerful near-ultraviolet (NUV) and visible imager onboard the Chinese Space Station Survey Telescope (CSST). The MCI provides three imaging channels, which are the NUV channel, the Blue channel and the Red channel, with the wavelength range of 255-430 nm, 430-700 nm, and 700-1000 nm, respectively. MCI's three channels can target the same field simultaneously, which is unique compared to other imagers onboard the Hubble Space Telescope (HST) or the James Webb Space Telescope (JWST). Each channel employs a CCD focal plane of 9216 x 9232 pixels and $\sim$7\arcmin.5 x 7\arcmin.5 field of view (FOV), which are about $\gtrsim 4$ times greater than the FOVs of HST imagers. The MCI's three channels feature unprecedented sensitivities and field of views complement the NUV and visible capabilities of the CSST for high-precision photometry and weak-signal detection, which would help build a new standard-star system and the deepest UV-Optical exposures for CSST. Rich filter sets of MCI would help explore other sciences such as local emission line mapping, high-z Ly$α$ emitters searching, etc. Here we present key design features, results of current ground tests, and suggested observing strategies of the MCI.

astro-ph.IM↗

arXiv2025

RedVTP: Training-Free Acceleration of Diffusion Vision-Language Models Inference via Masked Token-Guided Visual Token Pruning

Vision-Language Models (VLMs) have achieved remarkable progress in multimodal reasoning and generation, yet their high computational demands remain a major challenge. Diffusion Vision-Language Models (DVLMs) are particularly attractive because they enable parallel token decoding, but the large number of visual tokens still significantly hinders their inference efficiency. While visual token pruning has been extensively studied for autoregressive VLMs (AVLMs), it remains largely unexplored for DVLMs. In this work, we propose RedVTP, a response-driven visual token pruning strategy that leverages the inference dynamics of DVLMs. Our method estimates visual token importance using attention from the masked response tokens. Based on the observation that these importance scores remain consistent across steps, RedVTP prunes the less important visual tokens from the masked tokens after the first inference step, thereby maximizing inference efficiency. Experiments show that RedVTP improves token generation throughput of LLaDA-V and LaViDa by up to 186% and 28.05%, respectively, and reduces inference latency by up to 64.97% and 21.87%, without compromising-and in some cases improving-accuracy.

cs.CV↗

arXiv2025

Annular Channel Eigenmodes: A Physical-Layer Approach to Suppressing OAM Modal Crosstalk

Modal crosstalk is a fundamental limitation for orbital angular momentum (OAM)-based spatial-division multiplexing. Here, we introduce Annular Channel Eigenmodes (ACEs) - rigorously derived as the optimal band-limited solution for maximizing energy concentration within distinct annular channels. This approach reformulates the design as a Hermitian eigenvalue problem, efficiently yielding optimal beams that are physically isolated in space. Numerical simulations demonstrate that under identical conditions, conventional Gaussian-enveloped perfect optical vortices (POVs) exhibit an average modal crosstalk of -16 dB, whereas ACEs suppress crosstalk to nearly -30 dB. Moreover, the crosstalk suppression of ACEs continues to improve exponentially with increasing channel width, while that of POVs saturates at a fundamental limit. We experimentally generated ACEs and confirmed a 36% enhancement in energy confinement relative to POVs. ACEs thus provide a physically robust basis for high-fidelity, high-density OAM communications.

physics.optics↗

arXiv2025

HIVTP: A Training-Free Method to Improve VLMs Efficiency via Hierarchical Visual Token Pruning Using Middle-Layer-Based Importance Score

Vision-Language Models (VLMs) have shown strong capabilities on diverse multimodal tasks. However, the large number of visual tokens output by the vision encoder severely hinders inference efficiency, and prior studies have shown that many of these tokens are not important and can therefore be safely pruned. In this work, we propose HIVTP, a training-free method to improve VLMs efficiency via hierarchical visual token pruning using a novel middle-layer-based importance score. Specifically, we utilize attention maps extracted from the middle layers of the vision encoder, which better reflect fine-grained and object-level attention, to estimate visual token importance. Based on this, we propose a hierarchical visual token pruning method to retain both globally and locally important visual tokens. Specifically, we reshape the 1-D visual token sequence output by the vision encoder into a 2-D spatial layout. In the global retaining stage, we divide the image into regions and retain tokens with higher importance scores in each region; in the local retaining stage, we then divide the image into small windows and retain the most important token in each local window. Experimental results show that our proposed method, HIVTP, can reduce the time-to-first-token (TTFT) of LLaVA-v1.5-7B and LLaVA-Next-7B by up to 50.0% and 55.1%, respectively, and improve the token generation throughput by up to 60.9% and 47.3%, without sacrificing accuracy, and even achieving improvements on certain benchmarks. Compared with prior works, HIVTP achieves better accuracy while offering higher inference efficiency.

cs.CV↗

arXiv2025

IL-SLAM: Intelligent Line-assisted SLAM Based on Feature Awareness for Dynamic Environments

Visual Simultaneous Localization and Mapping (SLAM) plays a crucial role in autonomous systems. Traditional SLAM methods, based on static environment assumptions, struggle to handle complex dynamic environments. Recent dynamic SLAM systems employ geometric constraints and deep learning to remove dynamic features, yet this creates a new challenge: insufficient remaining point features for subsequent SLAM processes. Existing solutions address this by continuously introducing additional line and plane features to supplement point features, achieving robust tracking and pose estimation. However, current methods continuously introduce additional features regardless of necessity, causing two problems: unnecessary computational overhead and potential performance degradation from accumulated low-quality additional features and noise. To address these issues, this paper proposes a feature-aware mechanism that evaluates whether current features are adequate to determine if line feature support should be activated. This decision mechanism enables the system to introduce line features only when necessary, significantly reducing computational complexity of additional features while minimizing the introduction of low-quality features and noise. In subsequent processing, the introduced line features assist in obtaining better initial camera poses through tracking, local mapping, and loop closure, but are excluded from global optimization to avoid potential negative impacts from low-quality additional features in long-term process. Extensive experiments on TUM datasets demonstrate substantial improvements in both ATE and RPE metrics compared to ORB-SLAM3 baseline and superior performance over other dynamic SLAM and multi-feature methods.

cs.RO↗

arXiv2025

SR-SLAM: Scene-reliability Based RGB-D SLAM in Diverse Environments

Visual simultaneous localization and mapping (SLAM) plays a critical role in autonomous robotic systems, especially where accurate and reliable measurements are essential for navigation and sensing. In feature-based SLAM, the quantityand quality of extracted features significantly influence system performance. Due to the variations in feature quantity and quality across diverse environments, current approaches face two major challenges: (1) limited adaptability in dynamic feature culling and pose estimation, and (2) insufficient environmental awareness in assessment and optimization strategies. To address these issues, we propose SRR-SLAM, a scene-reliability based framework that enhances feature-based SLAM through environment-aware processing. Our method introduces a unified scene reliability assessment mechanism that incorporates multiple metrics and historical observations to guide system behavior. Based on this assessment, we develop: (i) adaptive dynamic region selection with flexible geometric constraints, (ii) depth-assisted self-adjusting clustering for efficient dynamic feature removal in high-dimensional settings, and (iii) reliability-aware pose refinement that dynamically integrates direct methods when features are insufficient. Furthermore, we propose (iv) reliability-based keyframe selection and a weighted optimization scheme to reduce computational overhead while improving estimation accuracy. Extensive experiments on public datasets and real world scenarios show that SRR-SLAM outperforms state-of-the-art dynamic SLAM methods, achieving up to 90% improvement in accuracy and robustness across diverse environments. These improvements directly contribute to enhanced measurement precision and reliability in autonomous robotic sensing systems.

cs.RO↗

arXiv2025

Adaptive Prior Scene-Object SLAM for Dynamic Environments

Visual Simultaneous Localization and Mapping (SLAM) plays a vital role in real-time localization for autonomous systems. However, traditional SLAM methods, which assume a static environment, often suffer from significant localization drift in dynamic scenarios. While recent advancements have improved SLAM performance in such environments, these systems still struggle with localization drift, particularly due to abrupt viewpoint changes and poorly characterized moving objects. In this paper, we propose a novel scene-object-based reliability assessment framework that comprehensively evaluates SLAM stability through both current frame quality metrics and scene changes relative to reliable reference frames. Furthermore, to tackle the lack of error correction mechanisms in existing systems when pose estimation becomes unreliable, we employ a pose refinement strategy that leverages information from reliable frames to optimize camera pose estimation, effectively mitigating the adverse effects of dynamic interference. Extensive experiments on the TUM RGB-D datasets demonstrate that our approach achieves substantial improvements in localization accuracy and system robustness under challenging dynamic scenarios.

cs.RO↗

arXiv2025

Re-Initialization Token Learning for Tool-Augmented Large Language Models

Large language models have demonstrated exceptional performance, yet struggle with complex tasks such as numerical reasoning, plan generation. Integrating external tools, such as calculators and databases, into large language models (LLMs) is crucial for enhancing problem-solving capabilities. Current methods assign a unique token to each tool, enabling LLMs to call tools through token prediction-similar to word generation. However, this approach fails to account for the relationship between tool and word tokens, limiting adaptability within pre-trained LLMs. To address this issue, we propose a novel token learning method that aligns tool tokens with the existing word embedding space from the perspective of initialization, thereby enhancing model performance. We begin by constructing prior token embeddings for each tool based on the tool's name or description, which are used to initialize and regularize the learnable tool token embeddings. This ensures the learned embeddings are well-aligned with the word token space, improving tool call accuracy. We evaluate the method on tasks such as numerical reasoning, knowledge-based question answering, and embodied plan generation using GSM8K-XL, FuncQA, KAMEL, and VirtualHome datasets. The results demonstrate clear improvements over recent baselines, including CoT, REACT, ICL, and ToolkenGPT, indicating that our approach effectively augments LLMs with tools through relevant tokens across diverse domains.

cs.CL↗

arXiv2025

A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment

The remarkable success of Large Language Models (LLMs) has illuminated a promising pathway toward achieving Artificial General Intelligence for both academic and industrial communities, owing to their unprecedented performance across various applications. As LLMs continue to gain prominence in both research and commercial domains, their security and safety implications have become a growing concern, not only for researchers and corporations but also for every nation. Currently, existing surveys on LLM safety primarily focus on specific stages of the LLM lifecycle, e.g., deployment phase or fine-tuning phase, lacking a comprehensive understanding of the entire "lifechain" of LLMs. To address this gap, this paper introduces, for the first time, the concept of "full-stack" safety to systematically consider safety issues throughout the entire process of LLM training, deployment, and eventual commercialization. Compared to the off-the-shelf LLM safety surveys, our work demonstrates several distinctive advantages: (I) Comprehensive Perspective. We define the complete LLM lifecycle as encompassing data preparation, pre-training, post-training, deployment and final commercialization. To our knowledge, this represents the first safety survey to encompass the entire lifecycle of LLMs. (II) Extensive Literature Support. Our research is grounded in an exhaustive review of over 800+ papers, ensuring comprehensive coverage and systematic organization of security issues within a more holistic understanding. (III) Unique Insights. Through systematic literature analysis, we have developed reliable roadmaps and perspectives for each chapter. Our work identifies promising research directions, including safety in data generation, alignment techniques, model editing, and LLM-based agent systems. These insights provide valuable guidance for researchers pursuing future work in this field.

cs.CR↗

arXiv2025

FPAN: Mitigating Replication in Diffusion Models through the Fine-Grained Probabilistic Addition of Noise to Token Embeddings

Diffusion models have demonstrated remarkable potential in generating high-quality images. However, their tendency to replicate training data raises serious privacy concerns, particularly when the training datasets contain sensitive or private information. Existing mitigation strategies primarily focus on reducing image duplication, modifying the cross-attention mechanism, and altering the denoising backbone architecture of diffusion models. Moreover, recent work has shown that adding a consistent small amount of noise to text embeddings can reduce replication to some degree. In this work, we begin by analyzing the impact of adding varying amounts of noise. Based on our analysis, we propose a fine-grained noise injection technique that probabilistically adds a larger amount of noise to token embeddings. We refer to our method as Fine-grained Probabilistic Addition of Noise (FPAN). Through our extensive experiments, we show that our proposed FPAN can reduce replication by an average of 28.78% compared to the baseline diffusion model without significantly impacting image quality, and outperforms the prior consistent-magnitude-noise-addition approach by 26.51%. Moreover, when combined with other existing mitigation methods, our FPAN approach can further reduce replication by up to 16.82% with similar, if not improved, image quality.

cs.CV↗

arXiv2025

Full space-time abrupt autofocusing spherical Airy wavepacket

The ability to precisely focus optical beams is crucial for numerous applications, yet conventional Gaussian beams exhibit slow intensity transitions near the focal point, limiting their effectiveness in scenarios requiring sharp focusing. In this work, the spherical Airy wavepacket, a three dimensional light field with an Airy function distribution in the radial direction in the full space time domain, is introduced and experimentally demonstrated. Leveraging the recently developed spatiotemporal hologram technique and an exponential polar coordinate transformation, spherical Airy wavepacket is sculpted to exhibit ultrafast autofocusing with a dramatically reduced depth of focus compared to conventional Gaussian beams and circular Airy beams. Experimental measurements confirm its nonlinear intensity surge and tight spatiotemporal confinement.

physics.optics↗

arXiv2025

Safe Human Robot Navigation in Warehouse Scenario

The integration of autonomous mobile robots (AMRs) in industrial environments, particularly warehouses, has revolutionized logistics and operational efficiency. However, ensuring the safety of human workers in dynamic, shared spaces remains a critical challenge. This work proposes a novel methodology that leverages control barrier functions (CBFs) to enhance safety in warehouse navigation. By integrating learning-based CBFs with the Open Robotics Middleware Framework (OpenRMF), the system achieves adaptive and safety-enhanced controls in multi-robot, multi-agent scenarios. Experiments conducted using various robot platforms demonstrate the efficacy of the proposed approach in avoiding static and dynamic obstacles, including human pedestrians. Our experiments evaluate different scenarios in which the number of robots, robot platforms, speed, and number of obstacles are varied, from which we achieve promising performance.

cs.RO↗

arXiv2025

Quality-focused Active Adversarial Policy for Safe Grasping in Human-Robot Interaction

Vision-guided robot grasping methods based on Deep Neural Networks (DNNs) have achieved remarkable success in handling unknown objects, attributable to their powerful generalizability. However, these methods with this generalizability tend to recognize the human hand and its adjacent objects as graspable targets, compromising safety during Human-Robot Interaction (HRI). In this work, we propose the Quality-focused Active Adversarial Policy (QFAAP) to solve this problem. Specifically, the first part is the Adversarial Quality Patch (AQP), wherein we design the adversarial quality patch loss and leverage the grasp dataset to optimize a patch with high quality scores. Next, we construct the Projected Quality Gradient Descent (PQGD) and integrate it with the AQP, which contains only the hand region within each real-time frame, endowing the AQP with fast adaptability to the human hand shape. Through AQP and PQGD, the hand can be actively adversarial with the surrounding objects, lowering their quality scores. Therefore, further setting the quality score of the hand to zero will reduce the grasping priority of both the hand and its adjacent objects, enabling the robot to grasp other objects away from the hand without emergency stops. We conduct extensive experiments on the benchmark datasets and a cobot, showing the effectiveness of QFAAP. Our code and demo videos are available here: https://github.com/clee-jaist/QFAAP.

cs.RO↗

arXiv2025

Fewer May Be Better: Enhancing Offline Reinforcement Learning with Reduced Dataset

Offline reinforcement learning (RL) represents a significant shift in RL research, allowing agents to learn from pre-collected datasets without further interaction with the environment. A key, yet underexplored, challenge in offline RL is selecting an optimal subset of the offline dataset that enhances both algorithm performance and training efficiency. Reducing dataset size can also reveal the minimal data requirements necessary for solving similar problems. In response to this challenge, we introduce ReDOR (Reduced Datasets for Offline RL), a method that frames dataset selection as a gradient approximation optimization problem. We demonstrate that the widely used actor-critic framework in RL can be reformulated as a submodular optimization objective, enabling efficient subset selection. To achieve this, we adapt orthogonal matching pursuit (OMP), incorporating several novel modifications tailored for offline RL. Our experimental results show that the data subsets identified by ReDOR not only boost algorithm performance but also do so with significantly lower computational complexity.

cs.LG↗

arXiv2024

Interpreting and Improving Attention From the Perspective of Large Kernel Convolution

Attention mechanisms have significantly advanced visual models by capturing global context effectively. However, their reliance on large-scale datasets and substantial computational resources poses challenges in data-scarce and resource-constrained scenarios. Moreover, traditional self-attention mechanisms lack inherent spatial inductive biases, making them suboptimal for modeling local features critical to tasks involving smaller datasets. In this work, we introduce Large Kernel Convolutional Attention (LKCA), a novel formulation that reinterprets attention operations as a single large-kernel convolution. This design unifies the strengths of convolutional architectures locality and translation invariance with the global context modeling capabilities of self-attention. By embedding these properties into a computationally efficient framework, LKCA addresses key limitations of traditional attention mechanisms. The proposed LKCA achieves competitive performance across various visual tasks, particularly in data-constrained settings. Experimental results on CIFAR-10, CIFAR-100, SVHN, and Tiny-ImageNet demonstrate its ability to excel in image classification, outperforming conventional attention mechanisms and vision transformers in compact model settings. These findings highlight the effectiveness of LKCA in bridging local and global feature modeling, offering a practical and robust solution for real-world applications with limited data and resources.

cs.CV↗

arXiv2024

LoyalDiffusion: A Diffusion Model Guarding Against Data Replication

Diffusion models have demonstrated significant potential in image generation. However, their ability to replicate training data presents a privacy risk, particularly when the training data includes confidential information. Existing mitigation strategies primarily focus on augmenting the training dataset, leaving the impact of diffusion model architecture under explored. In this paper, we address this gap by examining and mitigating the impact of the model structure, specifically the skip connections in the diffusion model's U-Net model. We first present our observation on a trade-off in the skip connections. While they enhance image generation quality, they also reinforce the memorization of training data, increasing the risk of replication. To address this, we propose a replication-aware U-Net (RAU-Net) architecture that incorporates information transfer blocks into skip connections that are less essential for image quality. Recognizing the potential impact of RAU-Net on generation quality, we further investigate and identify specific timesteps during which the impact on memorization is most pronounced. By applying RAU-Net selectively at these critical timesteps, we couple our novel diffusion model with a targeted training and inference strategy, forming a framework we refer to as LoyalDiffusion. Extensive experiments demonstrate that LoyalDiffusion outperforms the state-of-the-art replication mitigation method achieving a 48.63% reduction in replication while maintaining comparable image quality.

cs.CV↗

arXiv2024

A SAM-guided Two-stream Lightweight Model for Anomaly Detection

In industrial anomaly detection, model efficiency and mobile-friendliness become the primary concerns in real-world applications. Simultaneously, the impressive generalization capabilities of Segment Anything (SAM) have garnered broad academic attention, making it an ideal choice for localizing unseen anomalies and diverse real-world patterns. In this paper, considering these two critical factors, we propose a SAM-guided Two-stream Lightweight Model for unsupervised anomaly detection (STLM) that not only aligns with the two practical application requirements but also harnesses the robust generalization capabilities of SAM. We employ two lightweight image encoders, i.e., our two-stream lightweight module, guided by SAM's knowledge. To be specific, one stream is trained to generate discriminative and general feature representations in both normal and anomalous regions, while the other stream reconstructs the same images without anomalies, which effectively enhances the differentiation of two-stream representations when facing anomalous regions. Furthermore, we employ a shared mask decoder and a feature aggregation module to generate anomaly maps. Our experiments conducted on MVTec AD benchmark show that STLM, with about 16M parameters and achieving an inference time in 20ms, competes effectively with state-of-the-art methods in terms of performance, 98.26% on pixel-level AUC and 94.92% on PRO. We further experiment on more difficult datasets, e.g., VisA and DAGM, to demonstrate the effectiveness and generalizability of STLM.

cs.CV↗