arXiv subjects

Xiaolong Li

Publications and source records attributed to Xiaolong Li.

At least 55 records · Page 3Linked to original sources

arXiv2025

HY-Motion 1.0: Scaling Flow Matching Models for Text-To-Motion Generation

We present HY-Motion 1.0, a series of state-of-the-art, large-scale, motion generation models capable of generating 3D human motions from textual descriptions. HY-Motion 1.0 represents the first successful attempt to scale up Diffusion Transformer (DiT)-based flow matching models to the billion-parameter scale within the motion generation domain, delivering instruction-following capabilities that significantly outperform current open-source benchmarks. Uniquely, we introduce a comprehensive, full-stage training paradigm -- including large-scale pretraining on over 3,000 hours of motion data, high-quality fine-tuning on 400 hours of curated data, and reinforcement learning from both human feedback and reward models -- to ensure precise alignment with the text instruction and high motion quality. This framework is supported by our meticulous data processing pipeline, which performs rigorous motion cleaning and captioning. Consequently, our model achieves the most extensive coverage, spanning over 200 motion categories across 6 major classes. We release HY-Motion 1.0 to the open-source community to foster future research and accelerate the transition of 3D human motion generation models towards commercial maturity.

cs.CV↗

arXiv2025

PSA: Pyramid Sparse Attention for Efficient Video Understanding and Generation

Attention mechanisms are the core of foundation models, but their quadratic complexity remains a critical bottleneck for scaling. This challenge has driven the development of efficient attention mechanisms, with sparsity emerging as the dominant paradigm. Current methods typically retain or discard entire key-value blocks with binary masks, resulting in substantial information loss under high sparsity. To mitigate this gap, we present Pyramid Sparse Attention (PSA), a versatile module applicable to both video understanding and generation tasks. Instead of binary masking, PSA introduces multi-level pooled KV representations, enabling finer mask granularity. Specifically, each query block dynamically allocates lower pooling levels to critical KV blocks and higher levels to less important ones, creating an informative interpolation between full retention and complete pruning. This design, analogous to fixed-point quantization and classical feature pyramid networks in computer vision, effectively mitigates information loss while preserving computational efficiency under a low compute budget. It works with a native, hardware-friendly kernel that leverages decoupled block-tile design to ensure efficient execution. Across video understanding and generation benchmarks, PSA preserves contextual information and visual fidelity, consistently outperforming or achieving comparable performance over existing sparse attention baselines with superior efficiency-quality trade-offs. Our code and model weights are publicly available at: http://ziplab.co/PSA

cs.CV↗

arXiv2025

Histomorphology-Guided Prototypical Multi-Instance Learning for Breast Cancer WSI Classification

Histomorphology is crucial in cancer diagnosis. However, existing whole slide image (WSI) classification methods struggle to effectively incorporate histomorphology information, limiting their ability to capture key pathological features. Particularly when the number of instances within a bag is large and their features are complex, it becomes challenging to accurately identify instances decisive for the bag label, making these methods prone to interference from ambiguous instances. To address this limitation, we propose a novel Histomorphology-Guided Prototypical Multi-Instance Learning (HGPMIL) framework that explicitly learns histomorphology-guided prototypical representations by incorporating tumor cellularity, cellular morphology, and tissue architecture. Specifically, our approach consists of three key components: (1) estimating the importance of tumor-related histomorphology information at patch-level based on medical prior knowledge; (2) generating representative prototypes through histomorphology-prototypical clustering; and (3) enabling WSI classification through histomorphology-guided prototypical aggregation. HGPMIL adjusts the decision boundary by incorporating histomorphological importance to reduce instance label uncertainty, thereby reversely optimizing the bag-level boundary. Experimental results demonstrate its effectiveness, achieving high diagnostic accuracy for molecular subtyping, cancer subtyping and survival analysis. The code will be made available at https://github.com/Badgewho/HMDMIL.

cs.CV↗

arXiv2025

Too Good to be Bad: On the Failure of LLMs to Role-Play Villains

Large Language Models (LLMs) are increasingly tasked with creative generation, including the simulation of fictional characters. However, their ability to portray non-prosocial, antagonistic personas remains largely unexamined. We hypothesize that the safety alignment of modern LLMs creates a fundamental conflict with the task of authentically role-playing morally ambiguous or villainous characters. To investigate this, we introduce the Moral RolePlay benchmark, a new dataset featuring a four-level moral alignment scale and a balanced test set for rigorous evaluation. We task state-of-the-art LLMs with role-playing characters from moral paragons to pure villains. Our large-scale evaluation reveals a consistent, monotonic decline in role-playing fidelity as character morality decreases. We find that models struggle most with traits directly antithetical to safety principles, such as ``Deceitful'' and ``Manipulative'', often substituting nuanced malevolence with superficial aggression. Furthermore, we demonstrate that general chatbot proficiency is a poor predictor of villain role-playing ability, with highly safety-aligned models performing particularly poorly. Our work provides the first systematic evidence of this critical limitation, highlighting a key tension between model safety and creative fidelity. Our benchmark and findings pave the way for developing more nuanced, context-aware alignment methods.

cs.CL↗

arXiv2025

Gradient bounds for viscosity solutions to certain elliptic equations

Our principal object of study is the modulus of continuity of a periodic or uniformly vanishing function $ u: \mathbb{R} ^{n} \rightarrow \mathbb{R} $ which satisfies a degenerate elliptic equation $ F(x, u, \nabla u, D^{2} u) = 0 $ in the viscosity sense. The equations under consideration here have second-order terms of the form $ -{\rm Trace} \, (\mathcal{A} (\|\nabla u \|) \cdot D^{2} u) , $ where $ \mathcal{A} $ is an $ n\times n$ matrix which is symmetric and positive semi-definite. Following earlier work, \cite{Li21}, of the second author, which addressed the parabolic case, we identify a one-dimensional equation for which the modulus of continuity is a subsolution. In favorable cases, this one-dimensional operator can be used to derive a gradient bound on $u$ or to draw other conclusions about the nature of the solution.

math.AP↗

arXiv2025

Neural Network Based Framework for Passive Intermodulation Cancellation in MIMO Systems

Passive intermodulation (PIM) has emerged as a critical source of self-interference in modern MIMO-OFDM systems, especially under the stringent requirements of 5G and beyond. Conventional cancellation methods often rely on complex nonlinear models with limited scalability and high computational cost. In this work, we propose a lightweight deep learning framework for PIM cancellation that leverages depthwise separable convolutions and dilated convolutions to efficiently capture nonlinear dependencies across antennas and subcarriers. To further enhance convergence, we adopt a cyclic learning rate schedule and gradient clipping. In a controlled MIMO experimental setup, the method effectively suppresses third-order passive intermodulation (PIM) distortion, achieving up to 29dB of average power error (APE) with only 11k trainable parameters. These results highlight the potential of compact neural architectures for scalable interference mitigation in future wireless communication systems.

eess.SP↗

arXiv2025

Towards Reliable Pediatric Brain Tumor Segmentation: Task-Specific nnU-Net Enhancements

Accurate segmentation of pediatric brain tumors in multi-parametric magnetic resonance imaging (mpMRI) is critical for diagnosis, treatment planning, and monitoring, yet faces unique challenges due to limited data, high anatomical variability, and heterogeneous imaging across institutions. In this work, we present an advanced nnU-Net framework tailored for BraTS 2025 Task-6 (PED), the largest public dataset of pre-treatment pediatric high-grade gliomas. Our contributions include: (1) a widened residual encoder with squeeze-and-excitation (SE) attention; (2) 3D depthwise separable convolutions; (3) a specificity-driven regularization term; and (4) small-scale Gaussian weight initialization. We further refine predictions with two postprocessing steps. Our models achieved first place on the Task-6 validation leaderboard, attaining lesion-wise Dice scores of 0.759 (CC), 0.967 (ED), 0.826 (ET), 0.910 (NET), 0.928 (TC) and 0.928 (WT).

eess.IV↗

arXiv2025

New Sphere Theorems under Curvature Operator of the Second Kind

We investigate Riemannian manifolds $(M^n,g)$ whose curvature operator of the second kind $\mathring{R}$ satisfies the condition \begin{equation*} α^{-1} (λ_1 +\cdots +λ_α) > - θ\barλ, \end{equation*} where $λ_1 \leq \cdots \leq λ_{(n-1)(n+2)/2}$ are the eigenvalues of $\mathring{R}$, $\barλ$ is their average, and $θ> -1$. Under such conditions with optimal $θ$ depending on $n$ and $α$, we prove two differentiable sphere theorems in dimensions three and four, a homological sphere theorem in higher dimensions, and a curvature characterization of Kähler space forms. These results generalize recent works corresponding to $θ=0$ of Cao-Gursky-Tran, Nienhaus-Petersen-Wink, and the author. Moreover, examples are provided to demonstrate the sharpness of all results.

math.DG↗

arXiv2025

The Lighthouse of Language: Enhancing LLM Agents via Critique-Guided Improvement

Large language models (LLMs) have recently transformed from text-based assistants to autonomous agents capable of planning, reasoning, and iteratively improving their actions. While numerical reward signals and verifiers can effectively rank candidate actions, they often provide limited contextual guidance. In contrast, natural language feedback better aligns with the generative capabilities of LLMs, providing richer and more actionable suggestions. However, parsing and implementing this feedback effectively can be challenging for LLM-based agents. In this work, we introduce Critique-Guided Improvement (CGI), a novel two-player framework, comprising an actor model that explores an environment and a critic model that generates detailed nature language feedback. By training the critic to produce fine-grained assessments and actionable revisions, and the actor to utilize these critiques, our approach promotes more robust exploration of alternative strategies while avoiding local optima. Experiments in three interactive environments show that CGI outperforms existing baselines by a substantial margin. Notably, even a small critic model surpasses GPT-4 in feedback quality. The resulting actor achieves state-of-the-art performance, demonstrating the power of explicit iterative guidance to enhance decision-making in LLM-based agents.

cs.CL↗

arXiv2025

Micro-Act: Mitigating Knowledge Conflict in LLM-based RAG via Actionable Self-Reasoning

Retrieval-Augmented Generation (RAG) systems commonly suffer from Knowledge Conflicts, where retrieved external knowledge contradicts the inherent, parametric knowledge of large language models (LLMs). It adversely affects performance on downstream tasks such as question answering (QA). Existing approaches often attempt to mitigate conflicts by directly comparing two knowledge sources in a side-by-side manner, but this can overwhelm LLMs with extraneous or lengthy contexts, ultimately hindering their ability to identify and mitigate inconsistencies. To address this issue, we propose Micro-Act a framework with a hierarchical action space that automatically perceives context complexity and adaptively decomposes each knowledge source into a sequence of fine-grained comparisons. These comparisons are represented as actionable steps, enabling reasoning beyond the superficial context. Through extensive experiments on five benchmark datasets, Micro-Act consistently achieves significant increase in QA accuracy over state-of-the-art baselines across all 5 datasets and 3 conflict types, especially in temporal and semantic types where all baselines fail significantly. More importantly, Micro-Act exhibits robust performance on non-conflict questions simultaneously, highlighting its practical value in real-world RAG applications.

cs.CL↗

arXiv2025

Social Welfare Function Leaderboard: When LLM Agents Allocate Social Welfare

Large language models (LLMs) are increasingly entrusted with high-stakes decisions that affect human welfare. However, the principles and values that guide these models when distributing scarce societal resources remain largely unexamined. To address this, we introduce the Social Welfare Function (SWF) Benchmark, a dynamic simulation environment where an LLM acts as a sovereign allocator, distributing tasks to a heterogeneous community of recipients. The benchmark is designed to create a persistent trade-off between maximizing collective efficiency (measured by Return on Investment) and ensuring distributive fairness (measured by the Gini coefficient). We evaluate 20 state-of-the-art LLMs and present the first leaderboard for social welfare allocation. Our findings reveal three key insights: (i) A model's general conversational ability, as measured by popular leaderboards, is a poor predictor of its allocation skill. (ii) Most LLMs exhibit a strong default utilitarian orientation, prioritizing group productivity at the expense of severe inequality. (iii) Allocation strategies are highly vulnerable, easily perturbed by output-length constraints and social-influence framing. These results highlight the risks of deploying current LLMs as societal decision-makers and underscore the need for specialized benchmarks and targeted alignment for AI governance.

cs.CL↗

arXiv2025

The Hunger Game Debate: On the Emergence of Over-Competition in Multi-Agent Systems

LLM-based multi-agent systems demonstrate great potential for tackling complex problems, but how competition shapes their behavior remains underexplored. This paper investigates the over-competition in multi-agent debate, where agents under extreme pressure exhibit unreliable, harmful behaviors that undermine both collaboration and task performance. To study this phenomenon, we propose HATE, the Hunger Game Debate, a novel experimental framework that simulates debates under a zero-sum competition arena. Our experiments, conducted across a range of LLMs and tasks, reveal that competitive pressure significantly stimulates over-competition behaviors and degrades task performance, causing discussions to derail. We further explore the impact of environmental feedback by adding variants of judges, indicating that objective, task-focused feedback effectively mitigates the over-competition behaviors. We also probe the post-hoc kindness of LLMs and form a leaderboard to characterize top LLMs, providing insights for understanding and governing the emergent social dynamics of AI community.

cs.CL↗

arXiv2025

BatonVoice: An Operationalist Framework for Enhancing Controllable Speech Synthesis with Linguistic Intelligence from LLMs

The rise of Large Language Models (LLMs) is reshaping multimodel models, with speech synthesis being a prominent application. However, existing approaches often underutilize the linguistic intelligence of these models, typically failing to leverage their powerful instruction-following capabilities. This limitation hinders the model's ability to follow text instructions for controllable Text-to-Speech~(TTS). To address this, we propose a new paradigm inspired by ``operationalism'' that decouples instruction understanding from speech generation. We introduce BatonVoice, a framework where an LLM acts as a ``conductor'', understanding user instructions and generating a textual ``plan'' -- explicit vocal features (e.g., pitch, energy). A separate TTS model, the ``orchestra'', then generates the speech from these features. To realize this component, we develop BatonTTS, a TTS model trained specifically for this task. Our experiments demonstrate that BatonVoice achieves strong performance in controllable and emotional speech synthesis, outperforming strong open- and closed-source baselines. Notably, our approach enables remarkable zero-shot cross-lingual generalization, accurately applying feature control abilities to languages unseen during post-training. This demonstrates that objectifying speech into textual vocal features can more effectively unlock the linguistic intelligence of LLMs.

cs.CL↗

arXiv2025

BLADE: Block-Sparse Attention Meets Step Distillation for Efficient Video Generation

Diffusion Transformers currently lead the field in high-quality video generation, but their slow iterative denoising process and prohibitive quadratic attention costs for long sequences create significant inference bottlenecks. While both step distillation and sparse attention mechanisms have shown promise as independent acceleration strategies, effectively combining these approaches presents critical challenges -- training-free integration yields suboptimal results, while separately training sparse attention after step distillation requires prohibitively expensive high-quality video data. To overcome these limitations, we propose BLADE, an innovative data-free joint training framework that introduces: (1) an Adaptive Block-Sparse Attention (ASA) mechanism for dynamically generating content-aware sparsity masks to focus computation on salient spatiotemporal features, and (2) a sparsity-aware step distillation paradigm, built upon Trajectory Distribution Matching (TDM), directly incorporates sparsity into the distillation process rather than treating it as a separate compression step and features fast convergence. We validate BLADE on text-to-video models like CogVideoX-5B and Wan2.1-1.3B, and our framework demonstrates remarkable efficiency gains across different scales. On Wan2.1-1.3B, BLADE achieves a 14.10x end-to-end inference acceleration over a 50-step baseline. Moreover, on models such as CogVideoX-5B with short video sequence lengths, our framework delivers a robust 8.89x speedup. Crucially, the acceleration is accompanied by a consistent quality improvement. On the VBench-2.0 benchmark, BLADE boosts the score of CogVideoX-5B to 0.569 (from 0.534) and Wan2.1-1.3B to 0.570 (from 0.563), results that are further corroborated by superior ratings in human evaluations. Project is available at http://ziplab.co/BLADE-Homepage/.

cs.CV↗

arXiv2025

3D Aware Region Prompted Vision Language Model

We present Spatial Region 3D (SR-3D) aware vision-language model that connects single-view 2D images and multi-view 3D data through a shared visual token space. SR-3D supports flexible region prompting, allowing users to annotate regions with bounding boxes, segmentation masks on any frame, or directly in 3D, without the need for exhaustive multi-frame labeling. We achieve this by enriching 2D visual features with 3D positional embeddings, which allows the 3D model to draw upon strong 2D priors for more accurate spatial reasoning across frames, even when objects of interest do not co-occur within the same view. Extensive experiments on both general 2D vision language and specialized 3D spatial benchmarks demonstrate that SR-3D achieves state-of-the-art performance, underscoring its effectiveness for unifying 2D and 3D representation space on scene understanding. Moreover, we observe applicability to in-the-wild videos without sensory 3D inputs or ground-truth 3D annotations, where SR-3D accurately infers spatial relationships and metric measurements.

cs.CV↗

arXiv2025

Optimizing Convolution Direction and Template Selection for Difference Image Analysis

Difference image analysis (DIA) is a powerful tool for studying time-variable phenomena, and has been used by many time-domain surveys. Most DIA algorithms involve matching the spatially-varying PSF shape between science and template images, and then convolving that shape in one image to match the other. The wrong choice of which image to convolve can introduce one of the largest sources of artifacts in the final difference image. We introduce a quantitative metric to determine the optimal convolution direction that depends not only on the sharpness of the images measured by their FWHM, but also on their exposure depths. With this metric, the optimal convolution direction can be determined a priori, depending only on the FWHM and depth of the images. This not only simplifies the process, but also makes it more robust and less prone to creating sub-optimal difference images due to the wrong choice of the convolution direction. As an additional benefit, for a large set of images, we define a Figure-of-Merit based on this metric, which allows us to rank a list of images and determine the ones best suited to be used as templates, thus streamlining and automating the data reduction process.

astro-ph.IM↗

arXiv2025

The 9th AI City Challenge

The ninth AI City Challenge continues to advance real-world applications of computer vision and AI in transportation, industrial automation, and public safety. The 2025 edition featured four tracks and saw a 17% increase in participation, with 245 teams from 15 countries registered on the evaluation server. Public release of challenge datasets led to over 30,000 downloads to date. Track 1 focused on multi-class 3D multi-camera tracking, involving people, humanoids, autonomous mobile robots, and forklifts, using detailed calibration and 3D bounding box annotations. Track 2 tackled video question answering in traffic safety, with multi-camera incident understanding enriched by 3D gaze labels. Track 3 addressed fine-grained spatial reasoning in dynamic warehouse environments, requiring AI systems to interpret RGB-D inputs and answer spatial questions that combine perception, geometry, and language. Both Track 1 and Track 3 datasets were generated in NVIDIA Omniverse. Track 4 emphasized efficient road object detection from fisheye cameras, supporting lightweight, real-time deployment on edge devices. The evaluation framework enforced submission limits and used a partially held-out test set to ensure fair benchmarking. Final rankings were revealed after the competition concluded, fostering reproducibility and mitigating overfitting. Several teams achieved top-tier results, setting new benchmarks in multiple tasks.

cs.CV↗

arXiv2025

Takens' Last Problem and strong pluripotency

We consider the concept of strong pluripotency of dynamical systems for a hyperbolic invariant set, as introduced in [KNS]. To the best of our knowledge, for the whole hyperbolic invariant set, the existence of robust strongly pluripotent dynamical systems has not been proven in previous studies. In fact, there is an example of strongly pluripotent dynamical systems in [CV01], but its robustness has not been proven. On the other hand, robust strongly pluripotent dynamical systems for some proper subsets of hyperbolic sets had been found in [KS17, KNS]. In this paper, we provide a combinatorial way to recognize strongly pluripotent diffeomorphisms in a Newhouse domain and prove that they are $C^r$-robust, $2\leq r< \infty$. More precisely, we prove that there is a 2-dimensional diffeomorphism with a wild Smale horseshoe which has a $C^r$ neighborhood $\mathcal{U}_0$ where all elements are strongly pluripotent for the whole Smale horseshoe. Moreover, it follows from the result that any property, such as having a non-trivial physical measure supported by the Smale horseshoe or having historic behavior, is $C^r$-persistent relative to a dense subset of $\mathcal{U}_0$.

math.DS↗