paper_id uint32 0 5.29k | title stringlengths 14 183 | authors listlengths 1 36 | abstract large_stringlengths 246 3.59k | type stringclasses 3
values | arxiv_id stringlengths 10 10 ⌀ | github stringclasses 641
values | project_page stringclasses 244
values | space_ids listlengths 0 3 | model_ids listlengths 0 12 | dataset_ids listlengths 0 6 | embedding listlengths 768 768 |
|---|---|---|---|---|---|---|---|---|---|---|---|
0 | Feature-aware Modulation for Learning from Temporal Tabular Data | [
"Haorun Cai",
"Han-Jia Ye"
] | While tabular machine learning has achieved remarkable success, temporal distribution shifts pose significant challenges in real-world deployment, as the relationships between features and labels continuously evolve. Static models assume fixed mappings to ensure generalization, whereas adaptive models may overfit to transient patterns, creating a dilemma between robustness and adaptability.In this paper, we analyze key factors essential for constructing an effective dynamic mapping for temporal tabular data. We discover that evolving feature semantics—particularly objective and subjective meanings—introduce concept drift over time. Crucially, we identify that feature transformation strategies are able to mitigate discrepancies in feature representations across temporal stages.Motivated by these insights, we propose a feature-aware temporal modulation mechanism that conditions feature representations on temporal context, modulating statistical properties such as scale and skewness. By aligning feature semantics across time, our approach achieves a lightweight yet powerful adaptation, effectively balancing generalizability and adaptability.Benchmark evaluations validate the effectiveness of our method in handling temporal shifts in tabular data. | poster | null | null | null | [] | [] | [] | [
-0.027918590232729912,
-0.022952785715460777,
-0.00536078168079257,
0.022606397047638893,
0.04730525612831116,
0.012552446685731411,
0.02601015940308571,
0.011746042408049107,
-0.029478279873728752,
-0.020481018349528313,
-0.021653924137353897,
0.010324493981897831,
-0.054116588085889816,
... |
1 | Multimodal Tabular Reasoning with Privileged Structured Information | [
"Jun-Peng Jiang",
"Yu Xia",
"Hai-Long Sun",
"Shiyin Lu",
"Qingguo Chen",
"Weihua Luo",
"Kaifu Zhang",
"De-Chuan Zhan",
"Han-Jia Ye"
] | Tabular reasoning involves multi-step information extraction and logical inference over tabular data. While recent advances have leveraged large language models (LLMs) for reasoning over structured tables, such high-quality textual representations are often unavailable in real-world settings, where tables typically appear as images. In this paper, we tackle the task of tabular reasoning from table images, leveraging privileged structured information available during training to enhance multimodal large language models (MLLMs). The key challenges lie in the complexity of accurately aligning structured information with visual representations, and in effectively transferring structured reasoning skills to MLLMs despite the input modality gap. To address these, we introduce TabUlar Reasoning with Bridged infOrmation (Turbo), a new framework for multimodal tabular reasoning with privileged structured tables. Turbo benefits from a structure-aware reasoning trace generator based on DeepSeek-R1, contributing to high-quality modality-bridged data. On this basis, Turbo repeatedly generates and selects the advantageous reasoning paths, further enhancing the model's tabular reasoning ability. Experimental results demonstrate that, with limited ($9$k) data, Turbo achieves state-of-the-art performance ($+7.2\%$ vs. previous SOTA) across multiple datasets. | poster | 2506.04088 | null | null | [] | [] | [] | [
-0.024764923378825188,
-0.009755253791809082,
-0.0010533457389101386,
0.05830680578947067,
0.06296618282794952,
-0.008523516356945038,
0.005828986410051584,
0.0071451617404818535,
-0.030138159170746803,
-0.006433642003685236,
-0.009904402308166027,
0.030273418873548508,
-0.05234164744615555,... |
2 | Hawk: Leveraging Spatial Context for Faster Autoregressive Text-to-Image Generation | [
"Zhi-Kai Chen",
"Jun-Peng Jiang",
"Han-Jia Ye",
"De-Chuan Zhan"
] | Autoregressive (AR) image generation models can produce high-fidelity images but often struggle with slow inference due to their token-by-token, sequential decoding. Speculative decoding, which employs a draft model to approximate the AR model’s output, offers a promising way to reduce inference time. While this technique has been successfully applied to accelerate text-based AR models without sacrificing output quality, its application to image generation remains largely unexplored. Directly adapting this method to images is challenging because of the substantially larger sampling space, which complicates alignment between speculative and target model predictions, and the inadequate use of two-dimensional spatial information, which limits the exploitation of local image dependencies. To address these obstacles, we propose Spatial Speculative Decoding, a novel approach that leverages the inherent two-dimensional structure of images to guide a speculative model toward more accurate predictions and faster token generation. Experimental results on multiple text-to-image benchmarks demonstrate a 1.71× speedup over standard AR models, while preserving both image fidelity and diversity. | poster | null | null | null | [] | [] | [] | [
0.03323613107204437,
-0.003026977414265275,
-0.03981848806142807,
0.062294695526361465,
0.03403014317154884,
0.05531406030058861,
0.03570924326777458,
0.02824704721570015,
-0.0291270911693573,
-0.06182604283094406,
-0.03778502345085144,
0.006858844310045242,
-0.05914175137877464,
-0.003520... |
3 | AVR: Active Visual Reasoning for Multimodal Large Language Models in Physical Environments | [
"Weijie Zhou",
"Xuantang Xiong",
"Yi Peng",
"Manli Tao",
"Chaoyang Zhao",
"Honghui Dong",
"Ming Tang",
"Jinqiao Wang"
] | Visual reasoning in multimodal large language models (MLLMs) has primarily been studied in static, fully observable settings, limiting their effectiveness in real-world environments where information is often incomplete due to occlusion or limited field of view. Humans, in contrast, actively explore and interact with their environment—moving, examining, and manipulating objects—to gather information through a closed-loop process integrating perception, reasoning, and action. Inspired by this human capability, we introduce the Active Visual Reasoning (AVR) task, extending visual reasoning to partially observable, interactive environments. AVR necessitates agents to: (1) actively acquire information via sequential physical actions, (2) integrate observations across multiple steps for coherent reasoning, and (3) dynamically adjust decisions based on evolving visual feedback. To rigorously evaluate AVR, we introduce CLEVR-AVR, a simulation benchmark featuring multi-round interactive environments designed to assess both reasoning correctness and information-gathering efficiency. We present AVR-152k, a large-scale dataset offers rich Chain-of-Thought (CoT) annotations detailing iterative reasoning for uncertainty identification, action-conditioned information gain prediction, and information-maximizing action selection, crucial for training agents in a higher-order Markov Decision Process. Building on this, we develop PhysVLM-AVR, an MLLM achieving state-of-the-art performance on CLEVR-AVR, embodied reasoning (OpenEQA, RoboVQA), and passive visual reasoning (GeoMath, Geometry30K). Our analysis also reveals that current embodied MLLMs, despite detecting information incompleteness, struggle to actively acquire and integrate new information through interaction, highlighting a fundamental gap in active reasoning capabilities. | poster | null | null | null | [] | [] | [] | [
0.01666225679218769,
0.019161837175488472,
0.018184136599302292,
0.020116839557886124,
0.01080260518938303,
-0.0074893212877213955,
0.03461240977048874,
0.024095505475997925,
-0.061931390315294266,
-0.01807883195579052,
-0.032396890223026276,
0.03702402487397194,
-0.07656523585319519,
-0.0... |
4 | StelLA: Subspace Learning in Low-rank Adaptation using Stiefel Manifold | [
"Zhizhong Li",
"Sina Sajadmanesh",
"Jingtao Li",
"Lingjuan Lyu"
] | Low-rank adaptation (LoRA) has been widely adopted as a parameter-efficient technique for fine-tuning large-scale pre-trained models. However, it still lags behind full fine-tuning in performance, partly due to its insufficient exploitation of the geometric structure underlying low-rank manifolds. In this paper, we introduce a geometry-aware extension of LoRA that uses a three-factor decomposition $USV^\top$, separating the adapter's input and output subspaces $V$ and $U$ from the scaling component $S$, in the spirit of singular value decomposition (SVD). Our method constrains $U$ and $V$ to lie on the Stiefel manifold, ensuring their orthonormality throughout training. To optimize on the Stiefel manifold, we employ a flexible geometric optimization design that converts any Euclidean optimizer to a Riemannian optimizer via a modular interface. This enables principled and stable subspace learning while remaining compatible with existing fine-tuning pipelines. Empirical results across a wide range of downstream tasks, including commonsense reasoning, math and code generation, image classification, and image generation, demonstrate the superior performance of our approach against the vanilla LoRA and recent state-of-the-art variants. | spotlight | 2510.01938 | null | null | [] | [] | [] | [
-0.0012328630546107888,
-0.019064441323280334,
0.05699753016233444,
0.003122837282717228,
0.014146695844829082,
0.03768185153603554,
0.030706485733389854,
-0.01708410494029522,
-0.017843686044216156,
-0.03006056696176529,
-0.011502106674015522,
-0.011367151513695717,
-0.060943782329559326,
... |
5 | Continuous Subspace Optimization for Continual Learning | [
"Quan Cheng",
"Yuanyu Wan",
"Lingyu Wu",
"Chenping Hou",
"Lijun Zhang"
] | Continual learning aims to learn multiple tasks sequentially while preserving prior knowledge, but faces the challenge of catastrophic forgetting when acquiring new knowledge. Recently, approaches leveraging pre-trained models have gained increasing popularity to mitigate this issue, due to the strong generalization ability of foundation models. To adjust pre-trained models for new tasks, existing methods usually employ low-rank adaptation, which restricts parameter updates to a fixed low-rank subspace. However, constraining the optimization space inherently compromises the model's learning capacity, resulting in inferior performance. To address the limitation, we propose Continuous Subspace Optimization for Continual Learning (CoSO) to fine-tune the model in a series of subspaces rather than a single one. These sequential subspaces are dynamically determined through the singular value decomposition of gradients. CoSO updates the model by projecting gradients into these subspaces, ensuring memory-efficient optimization. To mitigate forgetting, the optimization subspaces of each task are set to be orthogonal to the historical task subspace. During task learning, CoSO maintains a task-specific component that captures the critical update directions associated with the current task. Upon completing a task, this component is used to update the historical task subspace, laying the groundwork for subsequent learning. Extensive experiments on multiple datasets demonstrate that CoSO significantly outperforms state-of-the-art methods, especially in challenging scenarios with long task sequences. | poster | 2505.11816 | null | null | [] | [] | [] | [
-0.025708384811878204,
-0.028254380449652672,
0.023082973435521126,
0.020732460543513298,
0.04052802175283432,
0.02129228413105011,
0.030985267832875252,
0.0076189348474144936,
-0.02262456715106964,
-0.026359064504504204,
-0.014774435199797153,
0.005049745552241802,
-0.08698197454214096,
-... |
6 | Point or Line? Using Line-based Representation for Panoptic Symbol Spotting in CAD Drawings | [
"Xingguang Wei",
"Haomin Wang",
"Shenglong Ye",
"Ruifeng Luo",
"Zhang",
"Lixin Gu",
"Jifeng Dai",
"Yu Qiao",
"Wenhai Wang",
"Hongjie Zhang"
] | We study the task of panoptic symbol spotting, which involves identifying both individual instances of countable \textit{things} and the semantic regions of uncountable \textit{stuff} in computer-aided design (CAD) drawings composed of vector graphical primitives.Existing methods typically rely on image rasterization, graph construction, or point-based representation, but these approaches often suffer from high computational costs, limited generality, and loss of geometric structural information. In this paper, we propose \textit{VecFormer}, a novel method that addresses these challenges through \textit{line-based representation} of primitives. This design preserves the geometric continuity of the original primitive, enabling more accurate shape representation while maintaining a computation-friendly structure, making it well-suited for vector graphic understanding tasks. To further enhance prediction reliability, we introduce a \textit{Branch Fusion Refinement} module that effectively integrates instance and semantic predictions, resolving their inconsistencies for more coherent panoptic outputs. Extensive experiments demonstrate that our method establishes a new state-of-the-art, achieving 91.1 PQ, with Stuff-PQ improved by 9.6 and 21.2 points over the second-best results under settings with and without prior information, respectively—highlighting the strong potential of line-based representation as a foundation for vector graphic understanding. | poster | 2505.23395 | null | null | [] | [] | [] | [
0.024368414655327797,
0.016579294577240944,
0.0037329329643398523,
0.01720212586224079,
0.032253995537757874,
0.04561635106801987,
0.011416948400437832,
0.024743089452385902,
-0.03564460575580597,
-0.07632939517498016,
-0.04257240891456604,
-0.017241783440113068,
-0.03736148774623871,
0.01... |
7 | HeroFilter: Adaptive Spectral Graph Filter for Varying Heterophilic Relations | [
"Shuaicheng Zhang",
"Haohui Wang",
"Junhong Lin",
"Xiaojie Guo",
"Yada Zhu",
"Si Zhang",
"Dongqi Fu",
"Dawei Zhou"
] | Graph heterophily, where connected nodes have different labels, has attracted significant interest recently. Most existing works adopt a simplified approach - using low-pass filters for homophilic graphs and high-pass filters for heterophilic graphs. However, we discover that the relationship between graph heterophily and spectral filters is more complex - the optimal filter response varies across frequency components and does not follow a strict monotonic correlation with heterophily degree. This finding challenges conventional fixed filter designs and suggests the need for adaptive filtering to preserve expressiveness in graph embeddings. Formally, natural questions arise: Given a heterophilic graph $\mathcal{G}$ , how and to what extent will the varying heterophily degree of $\mathcal{G}$ affect the performance of GNNs? How can we design adaptive filters to fit those varying heterophilic connections? Our theoretical analysis reveals that the average frequency response of GNNs and graph heterophily degree do not follow a strict monotonic correlation, necessitating adaptive graph filters to guarantee good generalization performance. Hence, we propose HeroFilter, a simple yet powerful GNN, which extracts information across the heterophily spectrum and combines salient representations through adaptive mixing. HeroFilter's superior performance achieves up to 9.2% accuracy improvement over leading baselines across homophilic and heterophilic graphs. | poster | 2510.10864 | null | null | [] | [] | [] | [
-0.005827364046126604,
-0.02612961269915104,
0.026585672050714493,
0.018297448754310608,
0.03338566794991493,
0.018741047009825706,
0.04036138206720352,
-0.002883843146264553,
-0.03569696098566055,
-0.0641101598739624,
0.0122721828520298,
-0.013883602805435658,
-0.10387987643480301,
-0.001... |
8 | Learning to Plan Like the Human Brain via Visuospatial Perception and Semantic-Episodic Synergistic Decision-Making | [
"Tianyuan Jia",
"Ziyu Li",
"Qing Li",
"Xiuxing Li",
"Xiang Li",
"Chen Wei",
"Li Yao",
"Xia Wu"
] | Motion planning in high-dimensional continuous spaces remains challenging due to complex environments and computational constraints. Although learning-based planners, especially graph neural network (GNN)-based, have significantly improved planning performance, they still struggle with inaccurate graph construction and limited structural reasoning, constraining search efficiency and path quality. The human brain exhibits efficient planning through a two-stage Perception-Decision model. First, egocentric spatial representations from visual and proprioceptive input are constructed, and then semantic–episodic synergy is leveraged to support decision-making in uncertainty scenarios. Inspired by this process, we propose NeuroMP, a brain-inspired planning framework that learns to plan like the human brain. NeuroMP integrates a Perceptive Segment Selector inspired by visuospatial perception to construct safer graphs, and a Global Alignment Heuristic guide search in weakly connected graphs by modeling semantic-episodic synergistic decision-making. Experimental results demonstrate that NeuroMP significantly outperforms existing planning methods in efficiency and quality while maintaining a high success rate. | poster | null | null | null | [] | [] | [] | [
-0.010123130865395069,
-0.0009883790044113994,
0.016352439299225807,
0.025908978655934334,
0.03345886245369911,
0.018588023260235786,
0.019428057596087456,
0.01947682723402977,
-0.029322246089577675,
-0.0647386759519577,
-0.03190242126584053,
-0.019755864515900612,
-0.048643894493579865,
-... |
9 | Cognitive Predictive Processing: A Human-like Framework for Adaptive Exploration in Open-World Reinforcement Learning | [
"boheng liu",
"Ziyu Li",
"Chenghua Duan",
"YuTian Liu",
"Zhuo Wang",
"Xiuxing Li",
"Qing Li",
"Xia Wu"
] | Open-world reinforcement learning challenges agents to develop intelligent behavior in vast exploration spaces. Recent approaches like LS-Imagine have advanced the field by extending imagination horizons through jumpy state transitions, yet remain limited by fixed exploration mechanisms and static jump thresholds that cannot adapt across changing task phases, resulting in inefficient exploration and lower completion rates. Humans demonstrate remarkable capabilities in open-world decision-making through a chain-like process of task decomposition, selective memory utilization, and adaptive uncertainty regulation. Inspired by human decision-making processes, we present Cognitive Predictive Processing (CPP), a novel framework that integrates three neurologically-inspired systems: a phase-adaptive cognitive controller that dynamically decomposes tasks into exploration, approach, and completion phases with adaptive parameters; a dual-memory integration system implementing dual-modal memory that balances immediate context with selective long-term storage; and an uncertainty-modulated prediction regulator that continuously updates environmental predictions to modulate exploration behavior. Comprehensive experiments in MineDojo demonstrate that these human-like decision-making strategies enhance performance over recent techniques, with success rates improving by an average of 4.6\% across resource collection tasks while reducing task completion steps by an average of 7.1\%. Our approach bridges cognitive neuroscience and reinforcement learning, excelling in complex scenarios that require sustained exploration and strategic adaptation while demonstrating how neural-inspired models can solve key challenges in open-world AI systems. Our main code has been anonymously uploaded to \url{https://anonymous.4open.science/r/CPP} without any author information. | poster | null | null | null | [] | [] | [] | [
-0.034635234624147415,
-0.0016626949654892087,
-0.004961226135492325,
0.02269046939909458,
0.06666268408298492,
0.01055567804723978,
0.013690098188817501,
0.02286916971206665,
-0.04776931181550026,
-0.04224323481321335,
-0.036726631224155426,
-0.00000944065504882019,
-0.053728241473436356,
... |
10 | FlexWorld: Progressively Expanding 3D Scenes for Flexible-View Exploration | [
"Luxi Chen",
"Zihan Zhou",
"Min Zhao",
"Yikai Wang",
"Ge Zhang",
"Wenhao Huang",
"Hao Sun",
"Ji-Rong Wen",
"Chongxuan LI"
] | Generating flexible-view 3D scenes, including 360° rotation and zooming, from single images is challenging due to a lack of 3D data. To this end, we introduce FlexWorld, a novel framework that progressively constructs a persistent 3D Gaussian splatting representation by synthesizing and integrating new 3D content. To handle novel view synthesis under large camera variations, we leverage an advanced pre-trained video model fine-tuned on accurate depth-estimated training pairs. By combining geometry-aware scene integration and optimization, FlexWorld refines the scene representation, producing visually consistent 3D scenes with flexible viewpoints. Extensive experiments demonstrate the effectiveness of FlexWorld in generating high-quality novel view videos and flexible-view 3D scenes from single images, achieving superior visual quality under multiple popular metrics and datasets compared to existing state-of-the-art methods. Additionally, FlexWorld supports extrapolating from existing 3D scenes, further extending its applicability. Qualitatively, we highlight that FlexWorld can generate high-fidelity scenes that enable 360° rotations and zooming exploration. | poster | null | null | null | [] | [] | [] | [
0.04716541990637779,
-0.021494794636964798,
0.04015105217695236,
0.03994341567158699,
0.03121611662209034,
0.0005947565659880638,
0.018641533330082893,
-0.002087100874632597,
-0.035839829593896866,
-0.05062934383749962,
-0.037267062813043594,
-0.025506269186735153,
-0.06009384244680405,
0.... |
11 | Learning Efficient Fuse-and-Refine for Feed-Forward 3D Gaussian Splatting | [
"Yiming Wang",
"Lucy Chai",
"Xuan Luo",
"Michael Niemeyer",
"Manuel Lagunas",
"Stephen Lombardi",
"Siyu Tang",
"Tiancheng Sun"
] | Recent advances in feed-forward 3D Gaussian Splatting have led to rapid improvements in efficient scene reconstruction from sparse views. However, most existing approaches construct Gaussian primitives directly aligned with the pixels in one or more of the input images. This leads to redundancies in the representation when input views overlap and constrains the position of the primitives to lie along the input rays without full flexibility in 3D space. Moreover, these pixel-aligned approaches do not naturally generalize to dynamic scenes, where effectively leveraging temporal information requires resolving both redundant and newly appearing content across frames. To address these limitations, we introduce a novel Fuse-and-Refine module that enhances existing feed-forward models by merging and refining the primitives in a canonical 3D space. At the core of our method is an efficient hybrid Splat-Voxel representation – from an initial set of pixel-aligned Gaussian primitives, we aggregate local features into a coarse-to-fine voxel hierarchy, and then use a sparse voxel transformer to process these voxel features and generate refined Gaussian primitives. By fusing and refining an arbitrary number of inputs into a consistent set of primitives, our representation effectively reduces redundancy and naturally adapts to temporal frames, enabling history-aware online reconstruction of dynamic scenes. Trained on large-scale static scene datasets, our model learns an effective global strategy to process around 20k primitives within 15ms and significantly enhances reconstruction quality compared to pixel-aligned reconstruction approaches. Without additional training, our model generalizes to video by fusing primitives across time, yielding a more temporally coherent result compared to baseline methods with graceful handling of occluded content. Our approach achieves state-of-the-art performance in both static and streaming scene reconstructions while running at interactive rates (15 fps with 350ms delay) on a single H100 GPU. | poster | null | null | null | [] | [] | [] | [
0.013603112660348415,
-0.0028351773507893085,
0.017727820202708244,
0.05550007149577141,
-0.013056181371212006,
0.01843137852847576,
0.013580935075879097,
0.027547208592295647,
-0.027254393324255943,
-0.04762668535113335,
-0.003512332681566477,
-0.01749984174966812,
-0.0784812793135643,
-0... |
12 | Implicit Modeling for Transferability Estimation of Vision Foundation Models | [
"Yaoyan Zheng",
"Huiqun Wang",
"Nan Zhou",
"Di Huang"
] | Transferability estimation identifies the best pre-trained models for downstream tasks without incurring the high computational cost of full fine-tuning. This capability facilitates deployment and advances the pre-training and fine-tuning paradigm. However, existing methods often struggle to accurately assess transferability for emerging pre-trained models with diverse architectures, training strategies, and task alignments. In this work, we propose Implicit Transferability Modeling (ITM), a novel framework that implicitly models each model’s intrinsic transferability, coupled with a Divide-and-Conquer Variational Approximation (DVA) strategy to efficiently approximate embedding space evolution. This design enables generalization across a broader range of models and downstream tasks. Extensive experiments on a comprehensive benchmark—spanning fuller training regimes and a wider variety of model types—demonstrate that ITM consistently outperforms existing methods in terms of stability, effectiveness, and efficiency. | poster | null | null | null | [] | [] | [] | [
-0.0021013091318309307,
-0.020286036655306816,
0.018581781536340714,
-0.002414434915408492,
0.04262903705239296,
0.034830108284950256,
0.0008661780739203095,
0.02529756724834442,
0.0022537345066666603,
-0.03551207482814789,
-0.011790711432695389,
0.02360781840980053,
-0.046600960195064545,
... |
13 | Neptune-X: Active X-to-Maritime Generation for Universal Maritime Object Detection | [
"Yu Guo",
"Shengfeng He",
"Yuxu Lu",
"Haonan An",
"Yihang Tao",
"Huilin Zhu",
"Jingxian Liu",
"Yuguang Fang"
] | Maritime object detection is essential for navigation safety, surveillance, and autonomous operations, yet constrained by two key challenges: the scarcity of annotated maritime data and poor generalization across various maritime attributes (e.g., object category, viewpoint, location, and imaging environment). To address these challenges, we propose Neptune-X, a data-centric generative-selection framework that enhances training effectiveness by leveraging synthetic data generation with task-aware sample selection. From the generation perspective, we develop X-to-Maritime, a multi-modality-conditioned generative model that synthesizes diverse and realistic maritime scenes. A key component is the Bidirectional Object-Water Attention module, which captures boundary interactions between objects and their aquatic surroundings to improve visual fidelity. To further improve downstream tasking performance, we propose Attribute-correlated Active Sampling, which dynamically selects synthetic samples based on their task relevance. To support robust benchmarking, we construct the Maritime Generation Dataset, the first dataset tailored for generative maritime learning, encompassing a wide range of semantic conditions. Extensive experiments demonstrate that our approach sets a new benchmark in maritime scene synthesis, significantly improving detection accuracy, particularly in challenging and previously underrepresented settings. | spotlight | null | null | null | [] | [] | [] | [
-0.005446244031190872,
-0.0338769368827343,
0.010229011066257954,
0.03692344203591347,
0.025040315464138985,
0.01683378592133522,
0.015548412688076496,
0.01621306501328945,
-0.0472722165286541,
-0.062531478703022,
-0.05164249241352081,
0.01610024832189083,
-0.06659263372421265,
-0.00604741... |
14 | Enhancing Sample Selection Against Label Noise by Cutting Mislabeled Easy Examples | [
"Suqin Yuan",
"Lei Feng",
"Bo Han",
"Tongliang Liu"
] | Sample selection is a prevalent approach in learning with noisy labels, aiming to identify confident samples for training. Although existing sample selection methods have achieved decent results by reducing the noise rate of the selected subset, they often overlook that not all mislabeled examples harm the model's performance equally. In this paper, we demonstrate that mislabeled examples correctly predicted by the model early in the training process are particularly harmful to model performance. We refer to these examples as Mislabeled Easy Examples (MEEs). To address this, we propose Early Cutting, which introduces a recalibration step that employs the model's later training state to re-select the confident subset identified early in training, thereby avoiding misleading confidence from early learning and effectively filtering out MEEs. Experiments on the CIFAR, WebVision, and full ImageNet-1k datasets demonstrate that our method effectively improves sample selection and model performance by reducing MEEs. | poster | 2502.08227 | null | null | [] | [] | [] | [
0.009575547650456429,
-0.026457756757736206,
-0.022491248324513435,
0.07333584129810333,
0.028811519965529442,
0.023864492774009705,
0.00645837839692831,
-0.010092352516949177,
-0.01787714846432209,
-0.062031883746385574,
-0.01723402366042137,
0.02480975352227688,
-0.08082041144371033,
0.0... |
15 | SAM2Flow: Interactive Optical Flow Estimation with Dual Memory for in vivo Microcirculation Analysis | [
"Luojie Huang",
"Ryan Zhang",
"Marisa Morakis",
"Michaela Taylor-Williams",
"Gregory McKay",
"Nicholas Durr"
] | Analysis of noninvasive microvascular blood flow can improve the diagnosis, prognosis, and management of many medical conditions, including cardiovascular, peripheral vascular, and sickle cell disease. This paper introduces SAM2Flow, an interactive optical flow estimation model to analyze long Oblique Back-illumination Microscopy (OBM) videos of in vivo microvascular flow. Inspired by the Segment Anything Model (SAM2), SAM2Flow enables users to specify regions of interest through user prompts for focused flow estimation. SAM2Flow also incorporates a dual memory attention mechanism, comprising both motion and context memory, to achieve efficient and stable flow estimations over extended video sequences. According to our experiments, SAM2Flow achieves SOTA accuracy in flow estimation with a fast inference speed of over $20$ fps on $512\times512$ inputs. Based on the temporally robust flow estimation, SAM2Flow demonstrated superior performance in downstream physiological applications compared to existing models. The code and dataset will be published with this paper. | poster | null | null | null | [] | [] | [] | [
-0.008561966940760612,
-0.01288684643805027,
0.01529319304972887,
-0.010093235410749912,
0.035856276750564575,
0.030157342553138733,
0.04474960267543793,
-0.006444050930440426,
-0.030102530494332314,
-0.06336484104394913,
0.02033141627907753,
-0.0421224981546402,
-0.059059567749500275,
0.0... |
16 | FastLongSpeech: Enhancing Large Speech-Language Models for Efficient Long-Speech Processing | [
"Shoutao Guo",
"Shaolei Zhang",
"Qingkai Fang",
"Zhengrui Ma",
"Min Zhang",
"Yang Feng"
] | The rapid advancement of Large Language Models (LLMs) has spurred significant progress in Large Speech-Language Models (LSLMs), enhancing their capabilities in both speech understanding and generation. While existing LSLMs often concentrate on augmenting speech generation or tackling a diverse array of short-speech tasks, the efficient processing of long-form speech remains a critical yet underexplored challenge. This gap is primarily attributed to the scarcity of long-speech training datasets and the high computational costs associated with long sequences. To address these limitations, we introduce FastLongSpeech, a novel framework designed to extend LSLM capabilities for efficient long-speech processing without necessitating dedicated long-speech training data. FastLongSpeech incorporates an iterative fusion strategy that can compress excessively long-speech sequences into manageable lengths. To adapt LSLMs for long-speech inputs, it introduces a dynamic compression training approach, which exposes the model to short-speech sequences at varying compression ratios, thereby transferring the capabilities of LSLMs to long-speech tasks. To assess the long-speech capabilities of LSLMs, we develop a long-speech understanding benchmark called LongSpeech-Eval. Experiments show that our method exhibits strong performance in both long-speech and short-speech tasks, while greatly improving inference efficiency. | poster | 2507.14815 | null | null | [] | [] | [] | [
-0.042389433830976486,
-0.043570276349782944,
-0.03024815395474434,
0.008672677911818027,
0.02926986664533615,
0.041353825479745865,
0.007362333592027426,
0.019666144624352455,
-0.03321342170238495,
-0.030115775763988495,
-0.032431505620479584,
0.026699958369135857,
-0.06559665501117706,
0... |
17 | Towards Human-Like Language Comprehension: Incremental and Wrap-Up Based fMRI-to-Text Decoding | [
"Wentao Lu",
"Dong Nie",
"Pengcheng Xue",
"Zheng Cui",
"Piji Li",
"Daoqiang Zhang",
"Xuyun Wen"
] | Decoding natural language text from non-invasive brain signals, such as functional magnetic resonance imaging (fMRI), remains a central challenge in brain-computer interface research. While recent advances in large language models (LLMs) have enabled open-vocabulary fMRI-to-text decoding, existing frameworks typically process the entire fMRI sequence in a single step, leading to performance degradation when handling long input sequences due to memory overload and semantic drift. To address this limitation, we propose a brain-inspired sequential fMRI-to-text decoding framework that mimics the human cognitive strategy of segmented and inductive language processing. Specifically, we divide long fMRI time series into consecutive segments aligned with optimal language comprehension length. Each segment is decoded incrementally, followed by a wrap-up mechanism that summarizes the semantic content and incorporates it as prior knowledge into subsequent decoding steps. This sequence-wise approach alleviates memory burden and ensures semantic continuity across segments. In addition, we introduce a text-guided masking strategy integrated with a masked autoencoder (MAE) framework for fMRI representation learning. This method leverages attention distributions over key semantic tokens to selectively mask the corresponding fMRI time points, and employs MAE to guide the model toward focusing on neural activity at semantically salient moments, thereby enhancing the capability of fMRI embeddings to represent textual information. Experimental results on the Narratives dataset demonstrate that our method significantly outperforms state-of-the-art approaches, with performance gains increasing as decoding length grows. | spotlight | null | null | null | [] | [] | [] | [
0.004505779128521681,
0.014873943291604519,
0.011986485682427883,
0.030309420078992844,
0.05041330307722092,
0.008924427442252636,
0.06576685607433319,
0.02759792096912861,
-0.024740057066082954,
-0.018252847716212273,
-0.019188037142157555,
0.02179001085460186,
-0.04361065477132797,
-0.01... |
18 | Value Gradient Guidance for Flow Matching Alignment | [
"Zhen Liu",
"Tim Xiao",
"Carles Domingo i Enrich",
"Weiyang Liu",
"Dinghuai Zhang"
] | While methods exist for aligning flow matching models -- a popular and effective class of generative models -- with human preferences, existing approaches fail to achieve both adaptation efficiency and probabilistically sound prior preservation. In this work, we leverage the theory of optimal control and propose VGG-Flow, a gradient matching–based method for finetuning pretrained flow matching models. The key idea in this algorithm is that the optimal difference between the finetuned velocity field and the pretrained one should be matched with the gradient field of a value function. This method not only incorporates first-order information from the reward model but also benefits from heuristic initialization of the value function to enable fast adaptation. Empirically, we show on a popular text-to-image flow matching model, Stable Diffusion 3, that our method can finetune flow matching models under limited computational budgets while achieving effective and prior-preserving alignment. | poster | null | null | null | [] | [] | [] | [
0.008065875619649887,
0.0010956383775919676,
0.02060890384018421,
0.046491414308547974,
0.021539080888032913,
0.06301402300596237,
0.031315211206674576,
0.014054049737751484,
-0.016195736825466156,
-0.057491425424814224,
-0.02328377775847912,
0.0074149142019450665,
-0.08627532422542572,
-0... |
19 | Metric Automata Theory: A Theory of Recurrent Neural Networks | [
"Adam Dankowiakowski",
"Alessandro Ronca"
] | We propose Metric Automata Theory, an elegant generalisation of classic Automata Theory to continuous dynamical systems, that constitutes a unifying theory of all kinds of Recurrent Neural Networks (RNNs), including widely-adopted architectures such as xLSTM and State Space Models (SSMs). The theory allows one to analyse RNNs both in the finite and unbounded precision settings seamlessly, and to use the fundamental results of Automata Theory. It also provides a novel notion of robustness that guarantees numerical stability, contributes to stability of learning, and notably allows one to prove results for real-world finite-precision implementations while abstracting away the difficulties introduced by finite-precision arithmetic.We employ the theory to prove a comprehensive set of expressivity results for widely-adopted recurrent neural architectures, with a focus on robustness and finite-precision. Notably, we prove that xLSTM, based on non-linear recurrence, is capable of recognising all star-free regular languages robustly, and hence by our theory it can also do so under any floating-point implementation given sufficient precision. Furthermore, we prove that linear recurrence, characterising SSMs, is not sufficient to robustly recognise all star-free languages. This provides a potential explanation for why xLSTM shows superior performance to SSMs on several tasks, and it gives a novel perspective on the importance of non-linear recurrences.We provide an improved understanding of the capabilities of Mamba, a popular SSM model. We show that Mamba is not generally capable of recognising the star-free languages under finite-precision, which is seemingly in contrast with the existing expressivity results for SSMs and also with its good empirical performance. We clarify the picture, by showing that Mamba admits a piecewise-linearly separable state space that allows it to approximate star-free languages, with some length-generalisation abilities. At the same time, Mamba does not admit such state spaces for languages like Parity. This explains why empirically Mamba performs well on star-free languages, and fails on Parity. | poster | null | null | null | [] | [] | [] | [
-0.04389554634690285,
-0.008876636624336243,
-0.018900856375694275,
0.02502347156405449,
0.03696039319038391,
0.04330931976437569,
0.027054911479353905,
0.024692660197615623,
-0.04065030813217163,
-0.014665118418633938,
-0.0010871835984289646,
-0.0074773565866053104,
-0.07725901901721954,
... |
20 | Breaking the Gold Standard: Extracting Forgotten Data under Exact Unlearning in Large Language Models | [
"Xiaoyu Wu",
"Yifei Pang",
"Terrance Liu",
"Steven Wu"
] | Large language models are typically trained on datasets collected from the web, which may inadvertently contain harmful or sensitive personal information. To address growing privacy concerns, unlearning methods have been proposed to remove the influence of specific data from trained models. Of these, exact unlearning---which retrains the model from scratch without the target data---is widely regarded the gold standard, believed to be robust against privacy-related attacks. In this paper, we challenge this assumption by introducing a novel data extraction attack that compromises even exact unlearning. Our method leverages both the pre- and post-unlearning models: by guiding the post-unlearning model using signals from the pre-unlearning model, we uncover patterns that reflect the removed data distribution. Combining model guidance with a token filtering strategy, our attack significantly improves extraction success rates---doubling performance in some cases---across common benchmarks such as MUSE, TOFU, and WMDP. Furthermore, we demonstrate our attack's effectiveness on a simulated medical diagnosis dataset to highlight real-world privacy risks associated with exact unlearning. In light of our findings, which suggest that unlearning may, in a contradictory way, *increase* the risk of privacy leakage, we advocate for evaluation of unlearning methods to consider broader threat models that account not only for post-unlearning models but also for adversarial access to prior checkpoints. | poster | null | null | null | [] | [] | [] | [
-0.023247765377163887,
-0.01573389209806919,
-0.015253865160048008,
0.0399298220872879,
0.04051969200372696,
-0.034864794462919235,
0.04842705279588699,
-0.004889218136668205,
-0.024439139291644096,
-0.007335699629038572,
-0.020658187568187714,
0.020644918084144592,
-0.05923030897974968,
-... |
21 | Continuous Thought Machines | [
"Luke Darlow",
"Ciaran Regan",
"Sebastian Risi",
"Jeffrey Seely",
"Llion Jones"
] | Biological brains demonstrate complex neural activity, where neural dynamics are critical to how brains process information. Most artificial neural networks simplify neurons by abstracting away dynamics. We challenge that paradigm. By incorporating neuron-level processing and synchronization, we reintroduce neural timing as a foundational element. We present the Continuous Thought Machine (CTM), a model designed to leverage neural dynamics as its core representation. The CTM has two innovations: (1) neuron-level temporal processing, where each neuron uses unique weight parameters to process incoming histories; and (2) neural synchronization as a latent representation. The CTM aims to strike a balance between neuron abstractions and biological realism. It operates at a level of abstraction that effectively captures essential temporal dynamics while remaining computationally tractable. We demonstrate the CTM's performance and versatility across a range of tasks, including solving 2D mazes, ImageNet-1K classification, parity computation, and more. Beyond displaying rich internal representations and offering a natural avenue for interpretation owing to its internal process, the CTM is able to perform tasks that require complex sequential reasoning. The CTM can also leverage adaptive compute, where it can stop earlier for simpler tasks, or keep computing when faced with more challenging instances. The goal of this work is to share the CTM and its associated innovations, rather than pushing for new state-of-the-art results. To that end, we believe the CTM represents a significant step toward developing more biologically plausible and powerful artificial intelligence systems. We provide an interactive online demonstration: https://anon-ctm.github.io/ctm/. | spotlight | 2505.05522 | https://github.com/SakanaAI/continuous-thought-machines | https://pub.sakana.ai/ctm/ | [] | [
"SakanaAI/ctm-maze-large",
"SakanaAI/ctm-imagenet"
] | [] | [
-0.016560055315494537,
0.002056199125945568,
-0.04764549061655998,
0.037892311811447144,
0.031436797231435776,
0.028509194031357765,
0.013122142292559147,
0.04207702353596687,
-0.053796131163835526,
-0.03178872913122177,
-0.0049812076613307,
-0.03211522474884987,
-0.055334486067295074,
0.0... |
22 | Improved Robust Estimation for Erdős-Rényi Graphs: The Sparse Regime and Optimal Breakdown Point | [
"Hongjie Chen",
"Jingqiu Ding",
"Yiding Hua",
"Stefan Tiegel"
] | We study the problem of robustly estimating the edge density of Erdos Renyi random graphs $\mathbb{G}(n, d^\circ/n)$ when an adversary can arbitrarily add or remove edges incident to an $\eta$-fraction of the nodes.We develop the first polynomial-time algorithm for this problem that estimates $d^\circ$ up to an additive error $O\left({[\sqrt{\log(n) / n} + \eta\sqrt{\log(1/\eta)} ] \cdot \sqrt{d^\circ} + \eta \log(1/\eta)}\right)$.Our error guarantee matches information-theoretic lower bounds up to factors of $\log(1/\eta)$.Moreover, our estimator works for all $d^\circ \geq \Omega(1)$ and achieves optimal breakdown point $\eta = 1/2$.Previous algorithms [Acharya et al 2022, Chen et al 2024], including inefficient ones, incur significantly suboptimal errors.Furthermore, even admitting suboptimal error guarantees, only inefficient algorithms achieve optimal breakdown point.Our algorithm is based on the sum-of-squares (SoS) hierarchy.A key ingredient is to construct constant-degree SoS certificates for concentration of the number of edges incident to small sets in $\mathbb{G}(n, d^\circ/n)$.Crucially, we show that these certificates also exist in the sparse regime, when $d^\circ = o(\log n)$, a regime in which the performance of previous algorithms was significantly suboptimal. | poster | null | null | null | [] | [] | [] | [
-0.021418185904622078,
0.020643122494220734,
0.010181562975049019,
0.0673799067735672,
0.03474370762705803,
0.04266872629523277,
0.013473333790898323,
0.002291695214807987,
-0.0215834379196167,
-0.06423566490411758,
0.0073342942632734776,
-0.037686195224523544,
-0.07148151844739914,
-0.000... |
23 | Flash Invariant Point Attention | [
"Andrew Liu",
"Axel Elaldi",
"Nicholas Franklin",
"Nathan Russell",
"Gurinder Atwal",
"Yih-En Ban",
"Olivia Viessmann"
] | Invariant Point Attention (IPA) is a key algorithm for geometry-aware modeling in structural biology, central to many protein and RNA models. However, its quadratic complexity limits the input sequence length. We introduce FlashIPA, a factorized reformulation of IPA that leverages hardware-efficient FlashAttention to achieve linear scaling in GPU memory and wall-clock time with sequence length. FlashIPA matches or exceeds standard IPA performance while substantially reducing computational costs. FlashIPA extends training to previously unattainable lengths, and we demonstrate this by re-training generative models without length restrictions and generating structures of thousands of residues. FlashIPA is available at https://anonymous.4open.science/r/flash_ipa-07CE. | spotlight | 2505.11580 | https://github.com/flagshippioneering/flash_ipa | null | [] | [] | [] | [
-0.023857861757278442,
0.0005595344700850546,
0.0015354336937889457,
0.042476460337638855,
-0.0000262320063484367,
0.014384922571480274,
0.035084184259176254,
0.028604499995708466,
-0.03340443968772888,
-0.008650088682770729,
0.011629338376224041,
-0.030122926458716393,
-0.0893365740776062,
... |
24 | Factor Decorrelation Enhanced Data Removal from Deep Predictive Models | [
"Wenhao Yang",
"Lin Li",
"Xiaohui Tao",
"Kaize Shi"
] | The imperative of user privacy protection and regulatory compliance necessitates sensitive data removal in model training, yet this process often induces distributional shifts that undermine model performance-particularly in out-of-distribution (OOD) scenarios. We propose a novel data removal approach that enhances deep predictive models through factor decorrelation and loss perturbation. Our approach introduces: (1) a discriminative-preserving factor decorrelation module employing dynamic adaptive weight adjustment and iterative representation updating to reduce feature redundancy and minimize inter-feature correlations. (2) a smoothed data removal mechanism with loss perturbation that creates information-theoretic safeguards against data leakage during removal operations. Extensive experiments on five benchmark datasets show that our approach outperforms other baselines and consistently achieves high predictive accuracy and robustness even under significant distribution shifts. The results highlight its superior efficiency and adaptability in both in-distribution and out-of-distribution scenarios. | poster | 2509.23443 | null | null | [] | [] | [] | [
0.02274608053267002,
-0.010641638189554214,
0.015833018347620964,
0.0454503558576107,
0.055874601006507874,
0.03865772858262062,
0.007830072194337845,
-0.026608798652887344,
-0.010165085084736347,
-0.018727686256170273,
-0.016745993867516518,
0.03311285376548767,
-0.07353731989860535,
0.01... |
25 | Accurately Predicting Protein Mutational Effects via a Hierarchical Many-Body Attention Network | [
"Dahao Xu",
"Jiahua Rao",
"Mingming Zhu",
"Jixian Zhang",
"Wei Lu",
"Shuangjia Zheng",
"Yuedong Yang"
] | Predicting changes in binding free energy ($\Delta\Delta G$) is essential for understanding protein-protein interactions, which are critical in drug design and protein engineering. However, existing methods often rely on pre-trained knowledge and heuristic features, limiting their ability to accurately model complex mutation effects, particularly higher-order and many-body interactions.To address these challenges, we propose H3-DDG, a Hypergraph-driven Hierarchical network to capture Higher-order many-body interactions across multiple scales. By introducing a hierarchical communication mechanism, H3-DDG effectively models both local and global mutational effects.Experimental results demonstrate state-of-the-art performance on multiple benchmarks. On the SKEMPI v2 dataset, H3-DDG achieves a Pearson correlation of 0.75, improving multi-point mutations prediction by 12.10%. On the challenging BindingGYM dataset, it outperforms Prompt-DDG and BA-DDG by 62.61% and 34.26%, respectively.Ablation and efficiency analyses demonstrate its robustness and scalability, while a case study on SARS-CoV-2 antibodies highlights its practical value in improving binding affinity for therapeutic design. | poster | null | null | null | [] | [] | [] | [
-0.018976787105202675,
0.02713700383901596,
-0.0010753106325864792,
0.0373331643640995,
0.04151134565472603,
-0.029275888577103615,
0.026754064485430717,
0.008724065497517586,
0.025108005851507187,
-0.05635906010866165,
0.036438461393117905,
-0.0049826521426439285,
-0.08858741074800491,
0.... |
26 | RGNMR: A Gauss-Newton method for robust matrix completion with theoretical guarantees | [
"Eilon Vaknin Laufer",
"Boaz Nadler"
] | Recovering a low rank matrix from a subset of its entries, some of which may be corrupted, is known as the robust matrix completion (RMC) problem.Existing RMC methods have several limitations: they require a relatively large number of observed entries; they may fail under overparametrization, when their assumed rank is higher than the correct one;and many of them fail to recover even mildly ill-conditioned matrices; In this paper we propose a novel RMC method, denoted $\texttt{RGNMR}$, which overcomes these limitations. $\texttt{RGNMR}$ is a simple factorization-based iterative algorithm, which combines a Gauss–Newton linearization with removal of entries suspected to be outliers.On the theoretical front, we prove that under suitable assumptions, $\texttt{RGNMR}$ is guaranteed exact recovery of the underlying low rank matrix.Our theoretical results improve upon the best currently known for factorization-based methods. On the empirical front, we show via several simulationsthe advantages of $\texttt{RGNMR}$ over existing RMC methods, and in particular its ability to handle a small number of observed entries, overparameterization of the rank and ill-conditioned matrices. | poster | 2505.12919 | null | null | [] | [] | [] | [
-0.018467172980308533,
-0.034171491861343384,
0.05669599026441574,
0.04053051769733429,
0.011751183308660984,
0.038107894361019135,
0.016240539029240608,
-0.015069558285176754,
-0.04642661288380623,
-0.04414965212345123,
-0.033900484442710876,
0.007610174361616373,
-0.0402606837451458,
-0.... |
27 | Dimension-Reduction Attack! Video Generative Models are Experts on Controllable Image Synthesis | [
"Hengyuan Cao",
"Yutong Feng",
"Biao Gong",
"Yijing Tian",
"Yunhong Lu",
"Chuang Liu",
"Bin Wang"
] | Video generative models can be regarded as world simulators due to their ability to capture dynamic, continuous changes inherent in real-world environments. These models integrate high-dimensional information across visual, temporal, spatial, and causal dimensions, enabling predictions of subjects in various status. A natural and valuable research direction is to explore whether a fully trained video generative model in high-dimensional space can effectively support lower-dimensional tasks such as controllable image generation. In this work, we propose a paradigm for video-to-image knowledge compression and task adaptation, termed \textit{Dimension-Reduction Attack} (\texttt{DRA-Ctrl}), which utilizes the strengths of video models, including long-range context modeling and flatten full-attention, to perform various generation tasks. Specially, to address the challenging gap between continuous video frames and discrete image generation, we introduce a mixup-based transition strategy that ensures smooth adaptation. Moreover, we redesign the attention structure with a tailored masking mechanism to better align text prompts with image-level control. Experiments across diverse image generation tasks, such as subject-driven and spatially conditioned generation, show that repurposed video models outperform those trained directly on images. These results highlight the untapped potential of large-scale video generators for broader visual applications. \texttt{DRA-Ctrl} provides new insights into reusing resource-intensive video models and lays foundation for future unified generative models across visual modalities. | poster | 2505.23325 | https://github.com/Kunbyte-AI/DRA-Ctrl | https://dra-ctrl-2025.github.io/DRA-Ctrl/ | [] | [] | [] | [
0.017379892989993095,
0.0011276182485744357,
-0.004833580460399389,
0.06862328946590424,
0.014701868407428265,
0.025431782007217407,
0.04075174406170845,
0.009691987186670303,
-0.017824314534664154,
-0.052476271986961365,
-0.03189433366060257,
-0.02591409720480442,
-0.055936381220817566,
0... |
28 | Reconstruction and Secrecy under Approximate Distance Queries | [
"Shay Moran",
"Elizaveta Nesterova"
] | Consider the task of locating an unknown target point using approximate distance queries: in each round, a reconstructor selects a reference point and receives a noisy version of its distance to the target. This problem arises naturally in various contexts—from localization in GPS and sensor networks to privacy-aware data access—making it relevant from the perspective of both the reconstructor (seeking accurate recovery) and the responder (aiming to limit information disclosure, e.g., for privacy or security reasons). We study this reconstruction game through a learning-theoretic lens, focusing on the rate and limits of the best possible reconstruction error.Our first result provides a tight geometric characterization of the optimal error in terms of the Chebyshev radius, a classical concept from geometry. This characterization applies to all compact metric spaces (in fact, to all totally bounded spaces) and yields explicit formulas for natural subsets of the Euclidean metric. Our second result addresses the asymptotic behavior of reconstruction, distinguishing between pseudo-finite spaces, where the optimal error is attained after finitely many queries, and spaces where the approximation curve exhibits a nontrivial decay. We characterize pseudo-finiteness for convex subsets of Euclidean spaces. | spotlight | null | null | null | [] | [] | [] | [
-0.01972268335521221,
0.0008075626683421433,
-0.0063181668519973755,
0.06400077790021896,
0.04994373768568039,
0.025050802156329155,
0.0273312795907259,
-0.012924923561513424,
-0.02179945632815361,
-0.05543757975101471,
-0.027157602831721306,
-0.02024850621819496,
-0.04917158558964729,
-0.... |
29 | Multi-Agent Reinforcement Learning with Communication-Constrained Priors | [
"Guang Yang",
"Jingwen Qiao",
"Tianpei Yang",
"Yanqing Wu",
"Jing Huo",
"Yang Gao",
"Xingguo Chen"
] | Communication is one of the effective means to improve the learning of cooperative policy in multi-agent systems. However, in most real-world scenarios, lossy communication is a prevalent issue. Existing multi-agent reinforcement learning with communication, due to their limited scalability and robustness, struggles to apply to complex and dynamic real-world environments. To address these challenges, we propose a generalized communication-constrained model to uniformly characterize communication conditions across different scenarios. Based on this, we utilize it as a learning prior to distinguish between lossy and lossless messages for specific scenarios. Additionally, we decouple the impact of lossy and lossless messages on distributed decision-making, drawing on a dual mutual information estimatior, and introduce a communication-constrained multi-agent reinforcement learning framework, quantifying the impact of communication messages into the global reward. Finally, we validate the effectiveness of our approach across several communication-constrained benchmarks. | poster | null | null | null | [] | [] | [] | [
-0.04174251854419708,
-0.028381625190377235,
-0.0029291585087776184,
0.05062014237046242,
0.03067026287317276,
0.02387108840048313,
0.010090578347444534,
-0.0119417579844594,
-0.01987997442483902,
-0.06683829426765442,
-0.028780197724699974,
0.03319651633501053,
-0.06470341980457306,
-0.00... |
30 | OmniTry: Virtual Try-On Anything without Masks | [
"Yutong Feng",
"Linlin Zhang",
"Hengyuan Cao",
"Yiming Chen",
"Xiaoduan Feng",
"Jian Cao",
"Yuxiong Wu",
"Bin Wang"
] | Virtual Try-ON (VTON) is a practical and widely-applied task, for which most of existing works focus on clothes. This paper presents OmniTry, a unified framework that extends VTON beyond garment to encompass any wearable objects, e.g., jewelries and accessories, with mask-free setting for more practical application. When extending to various types of objects, data curation is challenging for obtaining paired images, i.e., the object image and the corresponding try-on result. To tackle this problem, we propose a two-staged pipeline: For the first stage, we leverage large-scale unpaired images, i.e., portraits with any wearable items, to train the model for mask-free localization. Specifically, we repurpose the inpainting model to automatically draw objects in suitable positions given an empty mask. For the second stage, the model is further fine-tuned with paired images to transfer the consistency of object appearance. We observed that the model after the first stage shows quick convergence even with few paired samples. OmniTry is evaluated on a comprehensive benchmark consisting of 12 common classes of wearable objects, with both in-shop and in-the-wild images. Experimental results suggest that OmniTry shows better performance on both object localization and ID-preservation compared with existing methods. The code, model weights, and evaluation benchmark of OmniTry will be made publicly available. | poster | 2508.13632 | https://github.com/Kunbyte-AI/OmniTry | https://omnitry.github.io/ | [] | [] | [] | [
0.03785350173711777,
-0.00951524917036295,
0.002902213716879487,
0.01515879575163126,
0.04697934910655022,
0.026169223710894585,
0.035791102796792984,
0.0058857230469584465,
-0.043220266699790955,
-0.05741751194000244,
-0.053626466542482376,
-0.02863203175365925,
-0.057997431606054306,
-0.... |
31 | RiboFlow: Conditional De Novo RNA Co-Design via Synergistic Flow Matching | [
"Runze Ma",
"Zhongyue Zhang",
"Zichen Wang",
"Chenqing Hua",
"Jiahua Rao",
"Zhuomin Zhou",
"Shuangjia Zheng"
] | Ribonucleic acid (RNA) binds to molecules to achieve specific biological functions. While generative models are advancing biomolecule design, existing methods for designing RNA that target specific ligands face limitations in capturing RNA’s conformational flexibility, ensuring structural validity, and overcoming data scarcity. To address these challenges, we introduce RiboFlow, a synergistic flow matching model to co-design RNA structures and sequences based on target molecules. By integrating RNA backbone frames, torsion angles, and sequence features in an unified architecture, RiboFlow explicitly models RNA’s dynamic conformations while enforcing sequence-structure consistency to improve validity. Additionally, we curate RiboBind, a large-scale dataset of RNA-molecule interactions, to resolve the scarcity of high-quality structural data. Extensive experiments reveal that RiboFlow not only outperforms state-of-the-art RNA design methods by a large margin but also showcases controllable capabilities for achieving high binding affinity to target ligands. Our work bridges critical gaps in controllable RNA design, offering a framework for structure-aware, data-efficient generation. | poster | 2503.17007 | null | null | [] | [] | [] | [
-0.01119601633399725,
-0.001660739304497838,
-0.0016323043964803219,
0.007249234709888697,
0.03784212842583656,
-0.020492147654294968,
0.01454995758831501,
0.020318076014518738,
0.020007913932204247,
-0.03839743137359619,
0.010534442029893398,
0.001838024822063744,
-0.08810370415449142,
-0... |
32 | FPSAttention: Training-Aware FP8 and Sparsity Co-Design for Fast Video Diffusion | [
"Akide Liu",
"Zeyu Zhang",
"Zhexin Li",
"Xuehai Bai",
"Yizeng Han",
"Jiasheng Tang",
"Yuanjie Xing",
"Jichao Wu",
"Mingyang Yang",
"Weihua Chen",
"Jiahao He",
"Yuanyu He",
"Fan Wang",
"Reza Haffari",
"Bohan Zhuang"
] | Diffusion generative models have become the standard for producing high-quality, coherent video content, yet their slow inference speeds and high computational demands hinder practical deployment. Although both quantization and sparsity can independently accelerate inference while maintaining generation quality, naively combining these techniques in existing training-free approaches leads to significant performance degradation, as they fail to achieve proper joint optimization.We introduce FPSAttention, a novel training-aware co-design of FP8 quantization and Sparsity for video generation, with a focus on the 3D bi-directional attention mechanism. Our approach features three key innovations: 1) A unified 3D tile-wise granularity that simultaneously supports both quantization and sparsity. 2) A denoising step-aware strategy that adapts to the noise schedule, addressing the strong correlation between quantization/sparsity errors and denoising steps. 3) A native, hardware-friendly kernel that leverages FlashAttention and is implemented with optimized Hopper architecture features, enabling highly efficient execution.Trained on Wan2.1's 1.3B and 14B models and evaluated on the vBench benchmark, FPSAttention achieves a 7.09$\times$ kernel speedup for attention operations and a 4.96$\times$ end-to-end speedup for video generation compared to the BF16 baseline at 720p resolution—without sacrificing generation quality. | spotlight | 2506.04648 | null | https://fps.ziplab.co | [] | [] | [] | [
0.0018466348992660642,
-0.03485565632581711,
0.028592664748430252,
0.06197380647063255,
0.031667862087488174,
0.0404166541993618,
-0.0024280149955302477,
0.004284585360437632,
-0.032966550439596176,
-0.05167587846517563,
0.016125276684761047,
-0.04718352481722832,
-0.03536335378885269,
0.0... |
33 | Error Broadcast and Decorrelation as a Potential Artificial and Natural Learning Mechanism | [
"Mete Erdogan",
"Cengiz Pehlevan",
"Alper Erdogan"
] | We introduce *Error Broadcast and Decorrelation* (EBD), a novel learning framework for neural networks that addresses credit assignment by directly broadcasting output errors to individual layers, circumventing weight transport of backpropagation. EBD is rigorously grounded in the stochastic orthogonality property of Minimum Mean Square Error estimators. This fundamental principle states that the error of an optimal estimator is orthogonal to functions of the input. Guided by this insight, EBD defines layerwise loss functions that directly penalize correlations between layer activations and output errors, thereby establishing a principled foundation for error broadcasting. This theoretically sound mechanism naturally leads to the experimentally observed three-factor learning rule and integrates with biologically plausible frameworks to enhance performance and plausibility. Numerical experiments demonstrate EBD’s competitive or better performance against other error-broadcast methods on benchmark datasets. Our findings establish EBD as an efficient, biologically plausible, and principled alternative for neural network training. | spotlight | 2504.11558 | null | null | [] | [] | [] | [
-0.000257263018283993,
-0.00033894149237312376,
-0.009439353831112385,
0.02510307915508747,
0.0344238243997097,
0.0408477708697319,
0.012601342052221298,
-0.016166409477591515,
-0.03345580771565437,
-0.0330776609480381,
0.004915659315884113,
0.03788412734866142,
-0.052377086132764816,
-0.0... |
34 | ZPressor: Bottleneck-Aware Compression for Scalable Feed-Forward 3DGS | [
"Weijie Wang",
"Donny Y. Chen",
"Zeyu Zhang",
"Duochao Shi",
"Akide Liu",
"Bohan Zhuang"
] | Feed-forward 3D Gaussian Splatting (3DGS) models have recently emerged as a promising solution for novel view synthesis, enabling one-pass inference without the need for per-scene 3DGS optimization. However, their scalability is fundamentally constrained by the limited capacity of their encoders, leading to degraded performance or excessive memory consumption as the number of input views increases. In this work, we analyze feed-forward 3DGS frameworks through the lens of the Information Bottleneck principle and introduce ZPressor, a lightweight architecture-agnostic module that enables efficient compression of multi-view inputs into a compact latent state $Z$ that retains essential scene information while discarding redundancy. Concretely, ZPressor enables existing feed-forward 3DGS models to scale to over 100 input views at 480P resolution on an 80GB GPU, by partitioning the views into anchor and support sets and using cross attention to compress the information from the support views into anchor views, forming the compressed latent state $Z$. We show that integrating ZPressor into several state-of-the-art feed-forward 3DGS models consistently improves performance under moderate input views and enhances robustness under dense view settings on two large-scale benchmarks DL3DV-10K and RealEstate10K. | poster | 2505.23734 | https://github.com/ziplab/ZPressor | https://lhmd.top/zpressor/ | [] | [
"lhmd/ZPressor"
] | [] | [
0.017050398513674736,
-0.021259082481265068,
0.021852167323231697,
0.03567755967378616,
0.00837372150272131,
0.059140171855688095,
-0.012493755668401718,
0.017280615866184235,
-0.02314772829413414,
-0.04827477037906647,
-0.012910827063024044,
-0.010304654948413372,
-0.0695808157324791,
0.0... |
35 | Can Dependencies Induced by LLM-Agent Workflows Be Trusted? | [
"Yu Yao",
"Yiliao (Lia) Song",
"Yian Xie",
"Mengdan Fan",
"Mingyu Guo",
"Tongliang Liu"
] | LLM-agent systems often decompose a high-level task objective into a subtask-dependency graph, assuming each subtask’s response is conditionally independent of others given its parent responses. However, we find the inaccessible ground-truth responses will violate this assumption during execution, leading to inter-agent misalignment: failures arise from breakdowns in inter-agent interaction and coordination during execution. Consequently, both quality and runtime efficiency degenerate significantly. Motivated by this finding, we propose SeqCV, a dynamic framework that enables reliable execution under violated conditional independence assumptions. In SeqCV, subtasks are executed sequentially, each conditioned on all prior responses and verified via consistency checks immediately after agents generate a short token sequence. At each checkpoint, the token sequence is considered reliable if it is common knowledge consistently supported across diverse models. An unreliable token sequence is discarded, triggering a recursive splitting mechanism to decompose the subtask into more manageable components. Despite the sequential nature, SeqCV avoids costly misalignment corrections and delivers higher effective throughput than parallel pipelines. On different tasks, SeqCV not only improves accuracy by up to 17%, but also reduces execution time by more than half over six commonly used benchmarking datasets. | poster | null | null | null | [] | [] | [] | [
0.0009081226889975369,
-0.020997678861021996,
-0.039357319474220276,
0.035264741629362106,
0.03767700120806694,
0.020308738574385643,
0.042193345725536346,
0.00182359479367733,
-0.009220586158335209,
-0.03305625915527344,
-0.019461708143353462,
0.025730900466442108,
-0.08396648615598679,
-... |
36 | Pareto Optimal Risk-Agnostic Distributional Bandits with Heavy-Tail Rewards | [
"Kyungjae Lee",
"Dohyeong Kim",
"Taehyun Cho",
"Chaeyeon Kim",
"Yunkyung Ko",
"Seungyub Han",
"Seokhun Ju",
"Dohyeok Lee",
"Sungbin Lim"
] | This paper addresses the problem of multi-risk agnostic multi-armed bandits in heavy-tailed reward settings. We propose a framework that leverages novel deviation inequalities for the $1$-Wasserstein distance to construct confidence intervals for Lipschitz risk measures. The distributional LCB (DistLCB) algorithm is introduced, which achieves asymptotic optimality by deriving the first lower bounds for risk-aware bandits with explicit sub-optimality gap dependencies.The DistLCB is further extended to multi-risk objectives, which enables Pareto-optimal solutions that consider multiple aspects of reward distributions.Additionally, we provide a regret analysis that includes both gap-dependent and gap-independent bounds for multi-risk settings. Experiments validate the effectiveness of the proposed methods in synthetic and real-world applications. | poster | null | null | null | [] | [] | [] | [
-0.004672978073358536,
-0.003856652183458209,
0.006999914068728685,
0.020410090684890747,
0.0574500747025013,
0.01327209360897541,
0.007479086518287659,
0.013501251116394997,
-0.009086652658879757,
-0.029348602518439293,
-0.009422561153769493,
0.007519585080444813,
-0.06633301079273224,
-0... |
37 | Isotropic Noise in Stochastic and Quantum Convex Optimization | [
"Annie Marsden",
"Liam O'Carroll",
"Aaron Sidford",
"Chenyi Zhang"
] | We consider the problem of minimizing a $d$-dimensional Lipschitz convex function using a stochastic gradient oracle. We introduce and motivate a setting where the noise of the stochastic gradient is \emph{isotropic} in that it is bounded in every direction with high probability. We then develop an algorithm for this setting which improves upon prior results by a factor of $d$ in certain regimes, and as a corollary, achieves a new state-of-the-art complexity for sub-exponential noise. We give matching lower bounds (up to polylogarithmic factors) for both results. Additionally, we develop an efficient \emph{quantum isotropifier}, a quantum algorithm which converts a variance-bounded quantum sampling oracle into one that outputs an unbiased estimate with isotropic error. Combining our results, we obtain improved dimension-dependent rates for quantum stochastic convex optimization. | poster | null | null | null | [] | [] | [] | [
-0.04432869702577591,
0.023145776242017746,
-0.001615740591660142,
0.03358942270278931,
0.03163676708936691,
0.018071316182613373,
0.046398986130952835,
0.01484900712966919,
-0.0006638233317062259,
-0.06185572221875191,
-0.0180413406342268,
-0.03092058375477791,
-0.05527720972895622,
-0.01... |
38 | MixPrompt: Efficient Mixed Prompting for Multimodal Semantic Segmentation | [
"Zhiwei Hao",
"Zhongyu Xiao",
"Jianyuan Guo",
"Li Shen",
"Yong Luo",
"Han Hu",
"Dan Zeng"
] | Recent advances in multimodal semantic segmentation show that incorporating auxiliary inputs—such as depth or thermal images—can significantly improve performance over single-modality (RGB-only) approaches. However, most existing solutions rely on parallel backbone networks and complex fusion modules, greatly increasing model size and computational demands. Inspired by prompt tuning in large language models, we introduce \textbf{MixPrompt}: a prompting-based framework that integrates auxiliary modalities into a pretrained RGB segmentation model without modifying its architecture. MixPrompt uses a lightweight prompting module to extract and fuse information from auxiliary inputs into the main RGB backbone. This module is initialized using the early layers of a pretrained RGB feature extractor, ensuring a strong starting point. At each backbone layer, MixPrompt aligns RGB and auxiliary features in multiple low-rank subspaces, maximizing information use with minimal parameter overhead. An information mixing scheme enables cross-subspace interaction for further performance gains. During training, only the prompting module and segmentation head are updated, keeping the RGB backbone frozen for parameter efficiency. Experiments across NYU Depth V2, SUN-RGBD, MFNet, and DELIVER datasets show that MixPrompt achieves improvements of 4.3, 1.1, 0.4, and 1.1 mIoU, respectively, over two-branch baselines, while using nearly half the parameters. MixPrompt also outperforms recent prompting-based methods under similar compute budgets. | poster | null | null | null | [] | [] | [] | [
0.0061883688904345036,
-0.05437666177749634,
0.016899172216653824,
0.03529397025704384,
0.006876384373754263,
0.03421681374311447,
0.01799248531460762,
-0.0018392059719190001,
-0.057757969945669174,
-0.05331740155816078,
-0.05782344192266464,
0.012580710463225842,
-0.04326419532299042,
-0.... |
39 | Balancing Gradient and Hessian Queries in Non-Convex Optimization | [
"Deeksha Adil",
"Brian Bullins",
"Aaron Sidford",
"Chenyi Zhang"
] | We develop optimization methods which offer new trade-offs between the number of gradient and Hessian computations needed to compute the critical point of a non-convex function. We provide a method that for a twice-differentiable $f\colon \mathbb{R}^d \rightarrow \mathbb{R}$ with $L_2$-Lipschitz Hessian, and input initial point with $\Delta$-bounded sub-optimality and sufficiently small $\epsilon > 0$ outputs an $\epsilon$-critical point, i.e., a point $x$ such that $\|\nabla f(x)\| \leq \epsilon$, using $\tilde{O}(\Delta L_2^{1/4} n_H^{-1/2}\epsilon^{-9/4})$ queries to a gradient oracle and $n_H$ queries to a Hessian oracle. As a consequence, we obtain an improved gradient query complexity of $\tilde{O}(d^{1/3}L_2^{1/2}\Delta\epsilon^{-3/2})$ in the case of bounded dimension and of $\tilde{O}(\Delta^{3/2} L_2^{3/4}\epsilon^{-9/4})$ in the case where we are allowed only a single Hessian query. We obtain these results through a more general algorithm which can handle approximate Hessian computations and recovers known prior state-of-the-art bounds of computing an $\epsilon$-critical point, under the additional assumption that $f$ has an $L_1$-Lipschitz gradient, with $O(\Delta L_2^{1/4}\epsilon^{-7/4})$-gradient queries. | poster | null | null | null | [] | [] | [] | [
-0.061114583164453506,
-0.014550202526152134,
0.013816682621836662,
0.028039509430527687,
0.029253218322992325,
0.0386703722178936,
0.016547629609704018,
0.004170584492385387,
-0.028935866430401802,
-0.04593437910079956,
-0.013342607766389847,
0.011824709363281727,
-0.05888187512755394,
-0... |
40 | SPRINT: Enabling Interleaved Planning and Parallelized Execution in Reasoning Models | [
"Emil Biju",
"Shayan Talaei",
"Zhemin Huang",
"Mohammadreza Pourreza",
"Azalia Mirhoseini",
"Amin Saberi"
] | Large reasoning models (LRMs) excel at complex reasoning tasks but typically generate lengthy sequential chains-of-thought, resulting in long inference times before arriving at the final answer. To address this challenge, we introduce SPRINT, a novel post-training and inference-time framework designed to enable LRMs to dynamically identify and exploit opportunities for parallelization during their reasoning process. SPRINT incorporates an innovative data curation pipeline that reorganizes natural language reasoning trajectories into structured rounds of long-horizon planning and parallel execution. By fine-tuning LRMs on a small amount of such curated data, the models learn to dynamically identify independent subtasks within extended reasoning processes and effectively execute them in parallel. Through extensive evaluations, we show that the models fine-tuned with the SPRINT framework match the performance of reasoning models on complex domains such as mathematics while generating up to 39% fewer sequential tokens on problems requiring more than 8000 output tokens. Finally, we observe consistent results transferred to two out-of-distribution tasks of GPQA and Countdown with up to 45% and 65% reduction in average sequential tokens for longer reasoning trajectories, while achieving the performance of the fine-tuned reasoning model. | poster | 2506.05745 | null | null | [] | [] | [] | [
-0.028151802718639374,
-0.04177388176321983,
-0.037660688161849976,
0.035710908472537994,
0.05979843810200691,
0.011893108487129211,
0.00021256160107441247,
-0.002549184951931238,
-0.0504995658993721,
0.004843750968575478,
-0.00965619832277298,
0.008969835937023163,
-0.03887787461280823,
-... |
41 | Visual Sync: Multi‑Camera Synchronization via Cross‑View Object Motion | [
"Shaowei Liu",
"David Yao",
"Saurabh Gupta",
"Shenlong Wang"
] | Today, people can easily record memorable moments, ranging from concerts, sports events, lectures, family gatherings, and birthday parties with multiple consumer cameras. However, synchronizing these cross‑camera streams remains challenging. Existing methods assume controlled settings, specific targets, manual correction, or costly hardware. We present Visual Sync, an optimization framework based on multi‑view dynamics that aligns unposed, unsynchronized videos at millisecond accuracy. Our key insight is that any moving 3D point, when co‑visible in two cameras, obeys epipolar constraints once properly synchronized. To exploit this, Visual Sync leverages off‑the‑shelf 3D reconstruction, feature matching, and dense tracking to extract tracklets, relative poses, and cross‑view correspondences. It then jointly minimizes the epipolar error to estimate each camera’s time offset. Experiments on four diverse, challenging datasets show that Visual Sync outperforms baseline methods, achieving an average synchronization error below 130 ms. | poster | null | null | null | [] | [] | [] | [
0.03236149996519089,
0.01120869442820549,
0.01934487745165825,
0.03665389493107796,
0.025404131039977074,
0.03338450938463211,
0.01928810402750969,
0.038232482969760895,
-0.06633184105157852,
-0.06325594335794449,
-0.010844952426850796,
-0.036370035260915756,
-0.07617215067148209,
-0.03175... |
42 | CoDA: Coordinated Diffusion Noise Optimization for Whole-Body Manipulation of Articulated Objects | [
"Huaijin Pi",
"Zhi Cen",
"Zhiyang Dou",
"Taku Komura"
] | Synthesizing whole-body manipulation of articulated objects, including body motion, hand motion, and object motion, is a critical yet challenging task with broad applications in virtual humans and robotics.The core challenges are twofold.First, achieving realistic whole-body motion requires tight coordination between the hands and the rest of the body, as their movements are interdependent during manipulation. Second, articulated object manipulation typically involves high degrees of freedom and demands higher precision, often requiring the fingers to be placed at specific regions to actuate movable parts.To address these challenges, we propose a novel coordinated diffusion noise optimization framework.Specifically, we perform noise-space optimization over three specialized diffusion models for the body, left hand, and right hand, each trained on its own motion dataset to improve generalization.Coordination naturally emerges through gradient flow along the human kinematic chain, allowing the global body posture to adapt in response to hand motion objectives with high fidelity.To further enhance precision in hand-object interaction, we adopt a unified representation based on basis point sets (BPS), where end-effector positions are encoded as distances to the same BPS used for object geometry.This unified representation captures fine-grained spatial relationships between the hand and articulated object parts, and the resulting trajectories serve as targets to guide the optimization of diffusion noise, producing highly accurate interaction motion.We conduct extensive experiments demonstrating that our method outperforms existing approaches in motion quality and physical plausibility, and enables various capabilities such as object pose control, simultaneous walking and manipulation, and whole-body generation from hand-only data.The code will be released for reproducibility. | poster | 2505.21437 | https://github.com/phj128/CoDA | https://phj128.github.io/page/CoDA/index.html | [] | [] | [] | [
-0.0031036518048495054,
0.010619843378663063,
-0.03605518490076065,
0.020527957007288933,
0.04508993402123451,
0.043437566608190536,
0.02000703290104866,
0.001989541109651327,
-0.032605186104774475,
-0.07771720737218857,
-0.005116583779454231,
-0.03747495636343956,
-0.059963613748550415,
-... |
43 | Context-Aware Hierarchical Learning: A Two-Step Paradigm towards Safer LLMs | [
"Tengyun Ma",
"Damon Yao",
"Daojing He",
"Shihao Peng",
"YU LI",
"Shaohui Liu",
"Zhuotao Tian"
] | Large Language Models (LLMs) have emerged as powerful tools for diverse applications. However, their uniform token processing paradigm introduces critical vulnerabilities in instruction handling, particularly when exposed to adversarial scenarios. In this work, we identify and propose a novel class of vulnerabilities, termedTool-Completion Attack (TCA), which exploits function-calling mechanisms to subvert model behavior. To evaluate LLM robustness against such threats, we introduce the Tool-Completion Benchmark, a comprehensive security assessment framework, which reveals that even state-of-the-art models remain susceptible to TCA, with surprisingly high attack success rates. To address these vulnerabilities, we introduce Context-Aware Hierarchical Learning (CAHL), a sophisticated mechanism that dynamically equilibrates semantic comprehension with role-specific instruction constraints. CAHL leverages the contextual correlationsbetween different instruction segments to establish a robust, context-aware instruction hierarchy. Extensive experiments demonstrate that CAHL significantly enhances LLM robustness against both conventional attacks and the proposed TCA, exhibiting strong generalization capabilities in zero-shot evaluations while stillpreserving model performance on generic tasks. The proposed benchmark, code, and models willbe made publicly available. | poster | null | null | null | [] | [] | [] | [
-0.012016888707876205,
0.009073345921933651,
-0.03424068167805672,
0.046982377767562866,
0.022208403795957565,
-0.0006535480497404933,
0.040260523557662964,
0.014669499360024929,
-0.03271201252937317,
0.008761554025113583,
-0.022706113755702972,
0.017115361988544464,
-0.026187950745224953,
... |
44 | Vad-R1: Towards Video Anomaly Reasoning via Perception-to-Cognition Chain-of-Thought | [
"Chao Huang",
"Benfeng Wang",
"Jie Wen",
"Chengliang Liu",
"Wei Wang",
"Li Shen",
"Xiaochun Cao"
] | Recent advancements in reasoning capability of Multimodal Large Language Models (MLLMs) demonstrate its effectiveness in tackling complex visual tasks. However, existing MLLM-based Video Anomaly Detection (VAD) methods remain limited to shallow anomaly descriptions without deep reasoning. In this paper, we propose a new task named Video Anomaly Reasoning (VAR), which aims to enable deep analysis and understanding of anomalies in the video by requiring MLLMs to think explicitly before answering. To this end, we propose Vad-R1, an end-to-end MLLM-based framework for VAR. Specifically, we design a Perception-to-Cognition Chain-of-Thought (P2C-CoT) that simulates the human process of recognizing anomalies, guiding the MLLM to reason anomaly step-by-step. Based on the structured P2C-CoT, we construct Vad-Reasoning, a dedicated dataset for VAR. Furthermore, we propose an improved reinforcement learning algorithm AVA-GRPO, which explicitly incentivizes the anomaly reasoning capability of MLLMs through a self-verification mechanism with limited annotations. Experimental results demonstrate that Vad-R1 achieves superior performance, outperforming both open-source and proprietary models on VAD and VAR tasks. | poster | null | null | null | [] | [] | [] | [
0.025663523003458977,
0.003931472543627024,
0.020801087841391563,
0.043684493750333786,
0.03624672442674637,
-0.0036263868678361177,
0.03945501893758774,
-0.013499701395630836,
-0.043163541704416275,
-0.00822936650365591,
-0.017558835446834564,
0.038383398205041885,
-0.04882635176181793,
0... |
45 | MASTER: Enhancing Large Language Model via Multi-Agent Simulated Teaching | [
"Liang Yue",
"Yihong Tang",
"Kehai Chen",
"Jie Liu",
"Min Zhang"
] | Instruction fine-tuning is crucial in NLP tasks, enhancing pretrained models' instruction-following capabilities and task-specific performance. However, obtaining high-quality fine-tuning data for large models is challenging due to data collection difficulties and high production costs. To address this, we propose MASTER, a novel data augmentation method that enriches original data through interactions among multiple agents with varying cognitive levels. We simulate three pedagogically grounded teaching scenarios, leveraging multi-agent conversations to generate high-quality teacher-student interaction data. Utilizing MASTER, we construct BOOST-QA, a fine-tuning dataset augmented from existing datasets like Orca-Math-200k, ProcQA, and OpenHermes2.5. Experiments show that models fine-tuned with BOOST-QA perform excellently across multiple benchmarks, demonstrating strong multitask generalization. Notably, MASTER significantly improves models' reasoning abilities in complex tasks, providing valuable insights for future research. | poster | 2506.02689 | null | null | [] | [] | [] | [
-0.019295375794172287,
-0.030395545065402985,
-0.011135519482195377,
0.06032919883728027,
0.051379650831222534,
0.001007385435514152,
0.05105232074856758,
0.00957881286740303,
-0.0323137491941452,
-0.023708155378699303,
-0.0236616600304842,
0.056850578635931015,
-0.07461094856262207,
-0.01... |
46 | A Set of Generalized Components to Achieve Effective Poison-only Clean-label Backdoor Attacks with Collaborative Sample Selection and Triggers | [
"Zhixiao Wu",
"Yao Lu",
"Jie Wen",
"Hao Sun",
"Qi Zhou",
"Guangming Lu"
] | Poison-only Clean-label Backdoor Attacks (PCBAs) aim to covertly inject attacker-desired behavior into DNNs by merely poisoning the dataset without changing the labels. To effectively implant a backdoor, multiple triggers are proposed for various attack requirements of Attack Success Rate (ASR) and stealthiness. Additionally, sample selection enhances clean-label backdoor attacks' ASR by meticulously selecting "hard'' samples instead of random samples to poison. Current methods, however, 1) usually handle the sample selection and triggers in isolation, leading to severely limited improvements on both ASR and stealthiness. Consequently, attacks exhibit unsatisfactory performance on evaluation metrics when converted to PCBAs via a mere stacking of methods. Therefore, we seek to explore the bi-directional collaborative relations between the sample selection and triggers to address the above dilemma. 2) Since the strong specificity within triggers, the simple combination of sample selection and triggers fails to substantially enhance both evaluation metrics, with generalization preserved among various attacks. Therefore, we seek to propose a set of components to significantly improve both stealthiness and ASR based on the commonalities of attacks. Specifically, Component A ascertains two critical selection factors, and then makes them an appropriate combination based on the trigger scale to select more reasonable "hard'' samples for improving ASR. Component B is proposed to select samples with similarities to relevant trigger implanted samples to promote stealthiness. Component C reassigns trigger poisoning intensity on RGB colors through distinct sensitivity of the human visual system to RGB for higher ASR, with stealthiness ensured by sample selection including Component B. Furthermore, all components can be strategically integrated into diverse PCBAs, enabling tailored solutions that balance ASR and stealthiness enhancement for specific attack requirements. Extensive experiments demonstrate the superiority of our components in stealthiness, ASR, and generalization. Our code will be released as soon as possible. | poster | 2509.19947 | null | null | [] | [] | [] | [
-0.0036219267640262842,
-0.015948358923196793,
-0.019928330555558205,
0.048200417309999466,
0.03467373922467232,
0.0029588183388113976,
0.060068048536777496,
-0.010433814488351345,
-0.026482677087187767,
-0.03657764196395874,
-0.01874849759042263,
0.00736982561647892,
-0.06120599806308746,
... |
47 | Confidence-Aware With Prototype Alignment for Partial Multi-label Learning | [
"Weijun Lv",
"Yu Chen",
"Xuhuan Zhu",
"Jie Wen",
"Guoxu Zhou",
"Sixian Chan",
"Xiaozhao Fang"
] | Label prototype learning has emerged as an effective paradigm in Partial Multi-Label Learning (PML), providing a distinctive framework for modeling structured representations of label semantics while naturally filtering noise through prototype-based label confidence estimation. However, existing prototype-based methods face a critical limitation: class prototypes are the biased estimates due to noisy candidate labels, particularly when positive samples are scarce. To this end, we first propose a mutually class prototype alignment strategy bypassing noise interference by introducing two different transformation matrices, which makes the class prototypes learned by the fuzzy clustering and candidate label set mutually alignment for correcting themselves. Such alignment is also passed on to the fuzzy memberships label in turn. In addition, to eliminate noise interference in the candidate label set during the classifier learning, we use the learned permutation matrix to transform the fuzzy memberships label for learning a label enhancement indicator matrix accompanied by the candidate label set. This makes the label enhancement indicator matrix absolutely prevent the occurrence of numerical values located in non-label and simultaneously eliminate the introduction of incorrect label as much as possible. The resulting indicator matrix guides a robust multi-label classifier training process, jointly optimizing label confidence and classifier parameters. Extensive experiments demonstrate that our proposed model exhibits significant performance advantages over state-of-the-art PML approaches. | poster | null | null | null | [] | [] | [] | [
-0.028759630396962166,
-0.022512344643473625,
-0.03882458060979843,
0.04949561506509781,
0.013781799003481865,
0.02940058894455433,
0.03339296579360962,
-0.018238890916109085,
-0.01355264987796545,
-0.03174169734120369,
-0.01615791767835617,
0.024609897285699844,
-0.07459190487861633,
0.01... |
48 | Online locally differentially private conformal prediction via binary inquiries | [
"Qiangqiang Zhang",
"Chenfei Gu",
"Xinwei Feng",
"Jinhan Xie",
"Ting Li"
] | We propose an online conformal prediction framework under local differential privacy to address the emerging challenge of privacy-preserving uncertainty quantification in streaming data environments. Our method constructs dynamic, model-free prediction sets based on randomized binary inquiries, ensuring rigorous privacy protection without requiring access to raw data. Importantly, the proposed algorithm can be conducted in a one-pass online manner, leading to high computational efficiency and minimal storage requirements with $\mathcal{O}(1)$ space complexity, making it particularly suitable for real-time applications. The proposed framework is also broadly applicable to both regression and classification tasks, adapting flexibly to diverse predictive settings. We establish theoretical guarantees for long-run coverage at a target confidence level, ensuring statistical reliability under strict privacy constraints. Extensive empirical evaluations on both simulated and real-world datasets demonstrate that the proposed method delivers accurate, stable, and privacy-preserving predictions across a range of dynamic environments. | poster | null | null | null | [] | [] | [] | [
0.01616665907204151,
-0.012115166522562504,
0.009219417348504066,
0.05405449867248535,
0.056849319487810135,
0.010052134282886982,
0.03297341987490654,
-0.036852408200502396,
-0.018907275050878525,
-0.03093203715980053,
0.005755445919930935,
-0.028717877343297005,
-0.07988084107637405,
0.0... |
49 | Online robust locally differentially private learning for nonparametric regression | [
"Chenfei Gu",
"Qiangqiang Zhang",
"Ting Li",
"Jinhan Xie",
"Niansheng Tang"
] | The growing prevalence of streaming data and increasing concerns over data privacy pose significant challenges for traditional nonparametric regression methods, which are often ill-suited for real-time, privacy-aware learning. In this paper, we tackle these issuesby first proposing a novel one-pass online functional stochastic gradient descent algorithm that leverages the Huber loss (H-FSGD), to improve robustness against outliers and heavy-tailed errors in dynamic environments. To further accommodate privacy constraints, we introduce a locally differentially private extension, Private H-FSGD (PH-FSGD), designed to real-time, privacy-preserving estimation. Theoretically, we conduct a comprehensive non-asymptotic convergence analysis of the proposed estimators, establishing finite-sample guarantees and identifying optimal step size schedules that achieve optimal convergence rates. In particular, we provide practical insights into the impact of key hyperparameters, such as step size and privacy budget, on convergence behavior. Extensive experiments validate our theoretical findings, demonstrating that our methods achieve strong robustness and privacy protection without sacrificing efficiency. | poster | null | null | null | [] | [] | [] | [
-0.020248396322131157,
0.006029105745255947,
0.005734813399612904,
0.05205434560775757,
0.039663299918174744,
0.03457402437925339,
0.030081018805503845,
-0.0359211303293705,
-0.003884813981130719,
-0.032317619770765305,
0.012615101411938667,
-0.007137019652873278,
-0.052558910101652145,
0.... |
50 | Learning from Disjoint Views: A Contrastive Prototype Matching Network for Fully Incomplete Multi-View Clustering | [
"Yiming Wang",
"Qun Li",
"Dongxia Chang",
"Jie Wen",
"Hua Dai",
"Fu Xiao",
"Yao Zhao"
] | Multi-view clustering aims to enhance clustering performance by leveraging information from diverse sources. However, existing methods typically assume that instances are present in most or all views, which is impractical in real-world scenarios. This paper focuses on the understudied problem of fully incomplete multi-view clustering (FIMC), where each instance contains only a single view or cross-view correspondences are entirely unknown. To address this problem, we propose a Contrastive Prototype Matching Network (CPMN), which pioneers an explicit prototype matching strategy to effectively align cross-view instances through category-level correspondence. CPMN first employs a correspondence-free graph contrastive learning approach, leveraging mutual $k$-nearest neighbors (MNN) to uncover structural correlations and establish initial view-specific prototypes from entirely unpaired views. Building on the prototypes, we introduce a cross-view prototype graph matching stage to resolve category misalignment and forge a unified clustering structure. Finally, guided by this alignment, we devise a prototype-aware contrastive learning mechanism to promote semantic consistency, replacing the reliance on the initial MNN-based structural similarity. Extensive experiments on benchmark datasets demonstrate that our method significantly outperforms various baselines and ablation variants, validating its effectiveness. | poster | null | null | null | [] | [] | [] | [
0.005876854527741671,
-0.013286421075463295,
-0.00048601257731206715,
0.059144146740436554,
0.01756162941455841,
0.028318896889686584,
0.0048750340938568115,
-0.006273374892771244,
-0.0162566676735878,
-0.028231317177414894,
-0.011900125071406364,
0.00524180056527257,
-0.07736609131097794,
... |
51 | New Parallel and Streaming Algorithms for Directed Densest Subgraph | [
"Slobodan Mitrovic",
"Theodore Pan",
"Mahdi Qaempanah",
"Mohammad Amin Raeisi"
] | Finding dense subgraphs is a fundamental problem with applications to community detection, clustering, and data mining. Our work focuses on finding approximate densest subgraphs in directed graphs in computational models for processing massive data. We consider two such models: Massively Parallel Computation (MPC) and semi-streaming. We show how to find a $(2+\varepsilon)$-approximation in $\tilde{O}(\sqrt{\log n})$ MPC rounds with sublinear memory per machine. This improves the state-of-the-art results by Bahmani et al. (WAW 2014) and Mitrovic \& Pan (ICML 2024). Moreover, we show how to find an $O(\log n)$-approximation in a single pass in semi-streaming. This is in stark contrast to prior work, which implies $\tilde{\Omega}(n^{1/6})$-approximation for a single pass; a better approximation is known only for randomized streams (Mitrovi\'c \& Pan). We empirically evaluate our approaches in two ways. First, we illustrate that our single-pass semi-streaming algorithm performs much better than the theoretical guarantee. Specifically, its approximation on temporal datasets matches the $(2+\varepsilon)$-approximation of an $O(\log n)$-pass algorithm by Bahmani et al. (VLDB 2012). Second, we demonstrate that our MPC algorithm requires fewer rounds than prior work. | poster | 2509.21729 | null | null | [] | [] | [] | [
-0.013233757577836514,
-0.049270790070295334,
0.0035345065407454967,
0.0430755652487278,
0.038560446351766586,
0.023462513461709023,
0.026441382244229317,
0.016704196110367775,
-0.00525748822838068,
-0.05779717117547989,
0.007319603580981493,
-0.048799123615026474,
-0.08365750312805176,
0.... |
52 | Hierarchical Information Aggregation for Incomplete Multimodal Alzheimer's Disease Diagnosis | [
"Chengliang Liu",
"Que Yuanxi",
"Qihao Xu",
"Yabo Liu",
"Jie Wen",
"Jinghua Wang",
"Xiaoling Luo"
] | Alzheimer's Disease (AD) poses a significant health threat to the aging population, underscoring the critical need for early diagnosis to delay disease progression and improve patient quality of life. Recent advances in heterogeneous multimodal artificial intelligence (AI) have facilitated comprehensive joint diagnosis, yet practical clinical scenarios frequently encounter incomplete modalities due to factors like high acquisition costs or radiation risks. Moreover, traditional convolution-based architecture face inherent limitations in capturing long-range dependencies and handling heterogeneous medical data efficiently. To address these challenges, in our proposed heterogeneous multimodal diagnostic framework (HAD), we develop a multi-view Hilbert curve-based Mamba block and a hierarchical spatial feature extraction module to simultaneously capture local spatial features and global dependencies, effectively alleviating spatial discontinuities introduced by voxel serialization. Furthermore, to balance semantic consistency and modal specificity, we build a unified mutual information learning objective in the heterogeneous multimodal embedding space, which maintains effective learning of modality-specific information to avoid modality collapse caused by model preference. Extensive experiments demonstrate that our HAD significantly outperforms state-of-the-art methods in various modality-missing scenarios, providing an efficient and reliable solution for early-stage AD diagnosis. | poster | null | null | null | [] | [] | [] | [
-0.0181419737637043,
0.008326674811542034,
0.02492896281182766,
0.010062829591333866,
0.03200029209256172,
0.022991560399532318,
0.032194022089242935,
0.0014163692248985171,
-0.05136734992265701,
-0.05660361051559448,
-0.0037861361633986235,
0.011030643247067928,
-0.07305838912725449,
0.01... |
53 | NeuroH-TGL: Neuro-Heterogeneity Guided Temporal Graph Learning Strategy for Brain Disease Diagnosis | [
"Shengrong Li",
"Qi Zhu",
"Chunwei Tian",
"Xinyang Zhang",
"WEI SHAO",
"Jie Wen",
"Daoqiang Zhang"
] | Dynamic functional brain networks (DFBNs) are powerful tools in neuroscience research. Recent studies reveal that DFBNs contain heterogeneous neural nodes with more extensive connections and more drastic temporal changes, which play pivotal roles in coordinating the reorganization of the brain. Moreover, the spatio-temporal patterns of these nodes are modulated by the brain's historical states. However, existing methods not only ignore the spatio-temporal heterogeneity of neural nodes, but also fail to effectively encode the temporal propagation mechanism of heterogeneous activities. These limitations hinder the deep exploration of spatio-temporal relationships within DFBNs, preventing the capture of abnormal neural heterogeneity caused by brain diseases. To address these challenges, this paper propose a neuro-heterogeneity guided temporal graph learning strategy (NeuroH-TGL). Specifically, we first develop a spatio-temporal pattern decoupling module to disentangle DFBNs into topological consistency networks and temporal trend networks that align with the brain's operational mechanisms. Then, we introduce a heterogeneity mining module to identify pivotal heterogeneity nodes that drive brain reorganization from the two decoupled networks. Finally, we design temporal propagation graph convolution to simulate the influence of the historical states of heterogeneity nodes on the current topology, thereby flexibly extracting heterogeneous spatio-temporal information from the brain. Experiments show that our method surpasses several state-of-the-art methods, and can identify abnormal heterogeneous nodes caused by brain diseases. | poster | null | null | null | [] | [] | [] | [
-0.01269493903964758,
-0.0007733860402368009,
0.0010960986837744713,
0.019032886251807213,
0.04155169427394867,
0.0074620298109948635,
0.038897234946489334,
0.0063590300269424915,
-0.025801638141274452,
-0.04381495341658592,
0.0362679585814476,
-0.02568299137055874,
-0.048058364540338516,
... |
54 | Multi-Objective Hyperparameter Selection via Hypothesis Testing on Reliability Graphs | [
"Amirmohammad Farzaneh",
"Osvaldo Simeone"
] | The selection of hyperparameters, such as prompt templates in large language models (LLMs), must often strike a balance between reliability and cost. In many cases, structural relationships between the expected reliability levels of the hyperparameters can be inferred from prior information and held-out data -- e.g., longer prompt templates may be more detailed and thus more reliable. However, existing hyperparameter selection methods either do not provide formal reliability guarantees or are unable to incorporate structured knowledge in the hyperparameter space. This paper introduces reliability graph-based Pareto testing (RG-PT), a novel multi-objective hyperparameter selection framework that maintains formal reliability guarantees in terms of false discovery rate (FDR), while accounting for known relationships among hyperparameters via a directed acyclic graph. Edges in the graph reflect expected reliability and cost trade-offs among hyperparameters, which are inferred via the Bradley-Terry (BT) ranking model from prior information and held-out data. Experimental evaluations demonstrate that RG-PT significantly outperforms existing methods such as learn-then-test (LTT) and Pareto testing (PT) through a more efficient exploration of the hyperparameter space. | poster | 2501.13018 | null | null | [] | [] | [] | [
-0.028705749660730362,
0.009849018417298794,
0.004486558958888054,
0.07341478765010834,
0.014486103318631649,
0.06197265163064003,
0.044348180294036865,
-0.009934067726135254,
-0.007645355071872473,
-0.045513443648815155,
0.012137195095419884,
0.021447258070111275,
-0.050864510238170624,
-... |
55 | Backdoor Cleaning without External Guidance in MLLM Fine-tuning | [
"Xuankun Rong",
"Wenke Huang",
"Jian Liang",
"Jinhe Bi",
"Xun Xiao",
"Yiming Li",
"Bo Du",
"Mang Ye"
] | Multimodal Large Language Models (MLLMs) are increasingly deployed in fine-tuning-as-a-service (FTaaS) settings, where user-submitted datasets adapt general-purpose models to downstream tasks. This flexibility, however, introduces serious security risks, as malicious fine-tuning can implant backdoors into MLLMs with minimal effort. In this paper, we observe that backdoor triggers systematically disrupt cross-modal processing by causing abnormal attention concentration on non-semantic regions—a phenomenon we term **attention collapse**. Based on this insight, we propose **Believe Your Eyes (BYE)**, a data filtering framework that leverages attention entropy patterns as self-supervised signals to identify and filter backdoor samples. BYE operates via a three-stage pipeline: (1) extracting attention maps using the fine-tuned model, (2) computing entropy scores and profiling sensitive layers via bimodal separation, and (3) performing unsupervised clustering to remove suspicious samples. Unlike prior defenses, BYE equires no clean supervision, auxiliary labels, or model modifications. Extensive experiments across various datasets, models, and diverse trigger types validate BYE's effectiveness: it achieves near-zero attack success rates while maintaining clean-task performance, offering a robust and generalizable solution against backdoor threats in MLLMs. | poster | 2505.16916 | https://github.com/xuankunrong/bye | null | [] | [] | [] | [
-0.003589046886190772,
-0.001411370700225234,
0.019280729815363884,
0.0029271140228956938,
0.04393228143453598,
-0.00457167299464345,
0.053795404732227325,
0.001304445555433631,
-0.05662396550178528,
0.003308486193418503,
-0.02952934056520462,
0.029867663979530334,
-0.06723513454198837,
0.... |
56 | NormFit: A Lightweight Solution for Few-Shot Federated Learning with Non-IID Data | [
"Azadeh Motamedi",
"Jae-Mo Kang",
"Il-Min Kim"
] | Vision–Language Models (VLMs) have recently attracted considerable attention in Federated Learning (FL) due to their strong and robust performance. In particular, few-shot adaptation with pre-trained VLMs like CLIP enhances the performance of downstream tasks. However, existing methods still suffer from substantial communication overhead, high local computational demands, and suboptimal performance under non-IID user data. To simultaneously address all those limitations, we propose NormFit, a lightweight solution that selectively fine-tunes only a very small portion of the model parameters, specifically only the Pre-LayerNorm parameters of the vision encoder within a VLM. Overcoming the existing tradeoff between performance and communication/computation efficiency in few-shot FL, NormFit sets a new benchmark by simultaneously achieving superior accuracy and substantially reduced communication and computational demands. Theoretically, we show that NormFit yields a considerably smaller generalization gap compared to tuning all LayerNorm parameters. Importantly, NormFit can function effectively as a standalone solution or integrate seamlessly with existing few-shot fine-tuning methods to further enhance their performance. Notably, NormFit offers implementation simplicity, achieving these improvements without any algorithmic modifications, changes to the underlying model architecture, or the addition of external parameters. Code is provided as the supplementary material. | spotlight | null | null | null | [] | [] | [] | [
0.01733882911503315,
-0.06767328828573227,
0.014274779707193375,
0.04623536020517349,
0.026876412332057953,
0.014171460643410683,
0.030941206961870193,
-0.013910433277487755,
-0.018717430531978607,
-0.03437665104866028,
-0.019939150661230087,
0.01541103795170784,
-0.0667448565363884,
-0.01... |
57 | ECO: Evolving Core Knowledge for Efficient Transfer | [
"Fu Feng",
"Yucheng Xie",
"Ruixiao Shi",
"Jianlu Shen",
"Jingq Wang",
"Xin Geng"
] | Knowledge in modern neural networks is often entangled and structurally opaque, making current transfer methods—typically based on reusing entire parameter sets—inefficient and inflexible. Efforts to improve flexibility by reusing partial parameters frequently depend on handcrafted heuristics or rigid structural assumptions, which constrain generalization. In contrast, biological evolution enables efficient knowledge transfer by encoding only essential information into genes through iterative refinement under environmental pressure. Inspired by this principle, we propose \textbf{ECO}, a framework that \textbf{E}volves \textbf{CO}re knowledge into modular, reusable neural components—termed \textit{learngenes}—through similar evolutionary dynamics. To this end, we redefine learngenes as neural circuits and introduce Genetic Transfer Learning (GTL), a biologically inspired paradigm that establishes a genetic mechanism within neural networks. GTL simulates evolutionary processes by generating diverse network populations, selecting high-performing individuals, and transferring their learngenes to subsequent generations. Through iterative refinement, GTL enables learngenes to accumulate transferable, task-agnostic knowledge. Extensive experiments show that ECO achieves efficient initialization and strong generalization across diverse models and tasks, while significantly reducing computational and memory costs compared to conventional methods. | poster | null | null | null | [] | [] | [] | [
-0.006791040766984224,
-0.01177250687032938,
-0.004440586548298597,
0.021862458437681198,
0.051741644740104675,
0.027938442304730415,
0.01989133469760418,
0.012447614222764969,
-0.013297837227582932,
-0.023565860465168953,
0.009874069131910801,
0.012781466357409954,
-0.07693583518266678,
-... |
58 | Adaptive LoRA Experts Allocation and Selection for Federated Fine-Tuning | [
"Lei Wang",
"Jieming Bian",
"Letian Zhang",
"Jie Xu"
] | Large Language Models (LLMs) have demonstrated impressive capabilities across various tasks, but fine-tuning them for domain-specific applications often requires substantial domain-specific data that may be distributed across multiple organizations. Federated Learning (FL) offers a privacy-preserving solution, but faces challenges with computational constraints when applied to LLMs. Low-Rank Adaptation (LoRA) has emerged as a parameter-efficient fine-tuning approach, though a single LoRA module often struggles with heterogeneous data across diverse domains. This paper addresses two critical challenges in federated LoRA fine-tuning: 1. determining the optimal number and allocation of LoRA experts across heterogeneous clients, and 2. enabling clients to selectively utilize these experts based on their specific data characteristics. We propose FedLEASE (Federated adaptive LoRA Expert Allocation and SElection), a novel framework that adaptively clusters clients based on representation similarity to allocate and train domain-specific LoRA experts. It also introduces an adaptive top-$M$ Mixture-of-Experts mechanism that allows each client to select the optimal number of utilized experts. Our extensive experiments on diverse benchmark datasets demonstrate that FedLEASE significantly outperforms existing federated fine-tuning approaches in heterogeneous client settings while maintaining communication efficiency. | poster | 2509.15087 | null | null | [] | [] | [] | [
-0.02869234047830105,
-0.050214823335409164,
0.005425517912954092,
0.04308623820543289,
0.041183993220329285,
0.030644012615084648,
0.026226220652461052,
-0.013155006803572178,
-0.018448473885655403,
-0.016556553542613983,
-0.02382247895002365,
0.021786198019981384,
-0.05729927867650986,
0... |
59 | OmniVCus: Feedforward Subject-driven Video Customization with Multimodal Control Conditions | [
"Yuanhao Cai",
"HE Zhang",
"Xi Chen",
"Jinbo Xing",
"Yiwei Hu",
"Yuqian Zhou",
"Kai Zhang",
"Zhifei Zhang",
"Soo Ye Kim",
"Tianyu Wang",
"Yulun Zhang",
"Xiaokang Yang",
"Zhe Lin",
"Alan Yuille"
] | Existing feedforward subject-driven video customization methods mainly study single-subject scenarios due to the difficulty of constructing multi-subject training data pairs. Another challenging problem that how to use the signals such as depth, mask, camera, and text prompts to control and edit the subject in the customized video is still less explored. In this paper, we first propose a data construction pipeline, VideoCus-Factory, to produce training data pairs for multi-subject customization from raw videos without labels and control signals such as depth-to-video and mask-to-video pairs. Based on our constructed data, we develop an Image-Video Transfer Mixed (IVTM) training with image editing data to enable instructive editing for the subject in the customized video. Then we propose a diffusion Transformer framework, OmniVCus, with two embedding mechanisms, Lottery Embedding (LE) and Temporally Aligned Embedding (TAE). LE enables inference with more subjects by using the training subjects to activate more frame embeddings. TAE encourages the generation process to extract guidance from temporally aligned control signals by assigning the same frame embeddings to the control and noise tokens. Experiments demonstrate that our method significantly surpasses state-of-the-art methods in both quantitative and qualitative evaluations. | poster | 2506.23361 | https://github.com/caiyuanhao1998/Open-OmniVCus | https://caiyuanhao1998.github.io/project/OmniVCus/ | [] | [] | [] | [
0.03538781777024269,
-0.014539777301251888,
0.02360093593597412,
0.0518457256257534,
0.053519975394010544,
0.03444540873169899,
0.023586727678775787,
0.012247486971318722,
-0.01976233534514904,
-0.06285309046506882,
-0.018445448949933052,
-0.007769995369017124,
-0.053747814148664474,
-0.00... |
60 | LinEAS: End-to-end Learning of Activation Steering with a Distributional Loss | [
"Pau Rodriguez",
"Michal Klein",
"Eleonora Gualdoni",
"Valentino Maiorca",
"Arno Blaas",
"Luca Zappella",
"Marco Cuturi",
"Xavier Suau"
] | provide users with tools to explore style changes. Ideally, such mechanisms should require low volume of unpaired data (i.e., without explicit preference), and should be cheap, both at train and inference time, while preserving output quality. Recent research has shown that such mechanisms can be obtained by intervening exclusively on model activations, with the goal of correcting distributional differences between activations seen when using prompts from a source vs. a target set (e.g., toxic and non-toxic sentences). While cheap, these fast methods are inherently crude: their maps are tuned locally, not accounting for their impact on downstream layers, resulting in interventions that cause unintended shifts when used out-of-sample. We propose in this work linear end-to-end activation steering (LinEAS), an approach trained with a global loss that accounts simultaneously for all layer-wise distributional shifts. In addition to being more robust, the loss used to train LinEAS can be regularized with sparsifying norms, which can automatically carry out neuron selection. LinEAS only requires a handful of unpaired samples to be effective, and beats similar baselines on toxicity mitigation in language models, becoming competitive with oracle-dependent methods that have access to strong supervision. LinEAS is modality-agnostic and we empirically find that it outperforms existing activation steering methods at mitigating and including new concepts at the output of single-step text-to-image diffusion models. | poster | 2503.10679 | null | null | [] | [] | [] | [
-0.002331112278625369,
-0.015335300005972385,
-0.027253342792391777,
0.008779141120612621,
0.02663126401603222,
0.01658778451383114,
0.036294419318437576,
0.006024441681802273,
-0.009729166515171528,
-0.009852726012468338,
-0.025103885680437088,
0.022214792668819427,
-0.05116095393896103,
... |
61 | Geometry-Aware Collaborative Multi-Solutions Optimizer for Model Fine-Tuning with Parameter Efficiency | [
"Van-Anh Nguyen",
"Trung Le",
"Mehrtash Harandi",
"Ehsan Abbasnejad",
"Thanh-Toan Do",
"Dinh Phung"
] | We propose a framework grounded in gradient flow theory and informed by geometric structure that provides multiple diverse solutions for a given task, ensuring collaborative results that enhance performance and adaptability across different tasks. This framework enables flexibility, allowing for efficient task-specific fine-tuning while preserving the knowledge of the pre-trained foundation models. Extensive experiments across transfer learning, few-shot learning, and domain generalization show that our proposed approach consistently outperforms existing Bayesian methods, delivering strong performance with affordable computational overhead and offering a practical solution by updating only a small subset of parameters. | poster | null | null | null | [] | [] | [] | [
0.012545469217002392,
-0.0247485414147377,
0.025547362864017487,
0.03798225149512291,
0.009540080092847347,
0.06587615609169006,
0.01748024858534336,
-0.004145668353885412,
-0.007942241616547108,
-0.06644770503044128,
-0.011319808661937714,
-0.015492916107177734,
-0.06562022864818573,
-0.0... |
62 | FedEL: Federated Elastic Learning for Heterogeneous Devices | [
"Letian Zhang",
"Bo Chen",
"Jieming Bian",
"Lei Wang",
"Jie Xu"
] | Federated learning (FL) enables distributed devices to collaboratively train machine learning (ML) models while maintaining data privacy. However, the heterogeneous hardware capabilities of participating devices often result in significant training delays, as straggler clients with limited resources prolong the aggregation process. Existing solutions such as client selection, asynchronous FL, and partial training partially address these challenges but encounter issues such as reduced accuracy, stale updates, and compromised model performance due to inconsistent training contributions.To overcome these limitations, we propose FedEL, a federated elastic learning framework that enhances training efficiency while maintaining model accuracy. FedEL introduces a novel window-based training process, sliding the window to locate the training part of the modeland dynamically selecting important tensors for training within a coordinated runtime budget. This approach ensures progressive and balanced training across all clients, including stragglers. Additionally, FedEL employs a tensor importance adjustment module, harmonizing local and global tensor importance to mitigate biases caused by data heterogeneity. The experiment results shows that FedEL achieves up to 3.87× improvement in time-to-accuracy compared to baselines while maintaining or exceeding final test accuracy. | poster | 2509.16902 | null | null | [] | [] | [] | [
-0.017253980040550232,
-0.06470736861228943,
0.012990663759410381,
0.03954465687274933,
0.05173013359308243,
0.01588274911046028,
0.003291553584858775,
0.0017994551453739405,
-0.043141353875398636,
-0.05445457622408867,
-0.003571672597900033,
-0.0019156491616740823,
-0.039425354450941086,
... |
63 | TreeSynth: Synthesizing Diverse Data from Scratch via Tree-Guided Subspace Partitioning | [
"Sheng Wang",
"Pengan CHEN",
"Jingqi Zhou",
"Qintong Li",
"Jingwei Dong",
"Jiahui Gao",
"Boyang XUE",
"Jiyue Jiang",
"Lingpeng Kong",
"Chuan Wu"
] | Model customization necessitates high-quality and diverse datasets, but acquiring such data remains time-consuming and labor-intensive. Despite the great potential of large language models (LLMs) for data synthesis, current approaches are constrained by limited seed data, model biases and low-variation prompts, resulting in limited diversity and biased distribution with the increase of data scales.To tackle this challenge, we introduce TreeSynth, a tree-guided subspace-based data synthesis approach inspired by decision trees. It constructs a spatial partitioning tree to recursively divide a task-specific full data space (i.e., root node) into numerous atomic subspaces (i.e., leaf nodes) with mutually exclusive and exhaustive attributes to ensure both distinctiveness and comprehensiveness, before synthesizing samples within each atomic subspace. This globally divide-and-synthesize method finally collects subspace samples into a comprehensive dataset, effectively circumventing repetition and space collapse to ensure the diversity of large-scale data synthesis.Furthermore, the spatial partitioning tree enables sample allocation into atomic subspaces, allowing the re-balancing of existing datasets for more balanced and comprehensive distributions.Empirically, extensive experiments across diverse benchmarks consistently validates the superior data diversity, model performance, and robust scalability of TreeSynth compared to both human-crafted datasets and peer data synthesis methods, with the average performance gain reaching 10%.Besides, the consistent improvements of TreeSynth-balanced datasets highlight its efficacious application to redistribute existing datasets for more comprehensive coverage and the induced performance enhancement. The code is available in the anonymous repository via https://anonymous.4open.science/r/TreeSynth-EF04. | spotlight | 2503.17195 | https://github.com/cpa2001/TreeSynth | null | [] | [] | [] | [
-0.00042519107228145003,
-0.05035727098584175,
-0.0019922154024243355,
0.04350169375538826,
0.021539540961384773,
0.034891482442617416,
0.039044950157403946,
-0.006760315969586372,
-0.023873282596468925,
-0.0331762358546257,
-0.020345695316791534,
-0.0014024927513673902,
-0.08952596038579941... |
64 | Local-Global Coupling Spiking Graph Transformer for Brain Disorders Diagnosis from Two Perspectives | [
"Geng Zhang",
"Jiangrong Shen",
"Kaizhong Zheng",
"Liangjun Chen",
"Badong Chen"
] | Brain disorders have been consistently associated with abnormalities in specific brain regions or neural circuits. Identifying key brain regional activities and functional connectivity patterns is essential for discovering more precise neurobiological biomarkers. However, previous studies have primarily emphasized alterations in functional connectivity while overlooking abnormal neuronal population activity within brain regions. To bridge this gap, we propose a novel Local-Global Coupling Spiking Graph Transformer (LGC-SGT) that jointly models both inter-regional connectivity differences and deviations in neuronal population firing rates within brain regions, enabling a dual-perspective neuropathological analysis. The global pathway leverages spike-based computation in LGC-SGT to model biologically plausible aberrant neural firing dynamics, while the local pathway adaptively captures abnormal graph-based representations of brain connectivity learned by local plasticity in the liquid state machine module. Furthermore, we design a shortcut-enhanced output strategy in LGC-SGT with the hybrid loss function to suppress outlier interference caused by inter-individual and inter-center variability, enabling a more robust decision boundary. Extensive experiments on three brain disorder datasets demonstrate that our model consistently outperforms state-of-the-art graph methods in brain disorder diagnosis. Moreover, it facilitates the extraction of interpretable neurobiological biomarkers by jointly analyzing regional neural activity and functional connectivity, offering a more comprehensive framework for brain disorder understanding and diagnosis. | poster | null | null | null | [] | [] | [] | [
-0.020648270845413208,
-0.011780744418501854,
0.008928741328418255,
0.019519036635756493,
0.03507837653160095,
0.028640003874897957,
0.04256151244044304,
0.010836890898644924,
-0.02720564603805542,
-0.03429188206791878,
0.028346270322799683,
-0.010447405278682709,
-0.06329145282506943,
0.0... |
65 | RefLoRA: Refactored Low-Rank Adaptation for Efficient Fine-Tuning of Large Models | [
"Yilang Zhang",
"Bingcong Li",
"Georgios Giannakis"
] | Low-Rank Adaptation (LoRA) lowers the computational and memory overhead of fine-tuning large models by updating a low-dimensional subspace of the pre-trained weight matrix. Albeit efficient, LoRA exhibits suboptimal convergence and noticeable performance degradation, due to inconsistent and imbalanced weight updates induced by its nonunique low-rank factorizations. To overcome these limitations, this article identifies the optimal low-rank factorization per step that minimizes an upper bound on the loss. The resultant refactored low-rank adaptation (RefLoRA) method promotes a flatter loss landscape, along with consistent and balanced weight updates, thus speeding up stable convergence. Extensive experiments evaluate RefLoRA on natural language understanding, and commonsense reasoning tasks with popular large language models including DeBERTaV3, LLaMA-7B, LLaMA2-7B and LLaMA3-8B. The numerical tests corroborate that RefLoRA converges faster, outperforms various benchmarks, and enjoys negligible computational overhead compared to state-of-the-art LoRA variants. | poster | 2505.18877 | null | null | [] | [] | [] | [
-0.014658219181001186,
-0.033803414553403854,
-0.0009834003867581487,
0.03848820924758911,
0.018339920789003372,
0.038414839655160904,
0.012838461436331272,
0.005864795763045549,
-0.030769934877753258,
0.007347654551267624,
-0.006631434895098209,
0.03367539867758751,
-0.07495754212141037,
... |
66 | Recurrent Self-Attention Dynamics: An Energy-Agnostic Perspective from Jacobians | [
"Akiyoshi Tomihari",
"Ryo Karakida"
] | The theoretical understanding of self-attention (SA) has been steadily progressing. A prominent line of work studies a class of SA layers that admit an energy function decreased by state updates. While it provides valuable insights into inherent biases in signal propagation, it often relies on idealized assumptions or additional constraints not necessarily present in standard SA. Thus, to broaden our understanding, this work aims to relax these energy constraints and provide an energy-agnostic characterization of inference dynamics by dynamical systems analysis.In more detail, we first consider relaxing the symmetry and single-head constraints traditionally required in energy-based formulations. Next, to investigate more general SA architectures capable of oscillatory dynamics without necessarily admitting an energy function, we analyze the Jacobian matrix of the state. We reveal that normalization layers effectively normalize the Jacobian's complex eigenvalues, forcing the dynamics close to a critical state. This significantly enhances inference performance. Furthermore, we utilize the Jacobian perspective to develop regularization methods for training and a pseudo-energy for monitoring inference dynamics. | poster | 2505.19458 | null | null | [] | [] | [] | [
-0.03748534992337227,
-0.01629278063774109,
-0.0025501518975943327,
0.019434137269854546,
0.022137770429253578,
0.02256922796368599,
0.056500665843486786,
0.015863487496972084,
-0.050332602113485336,
-0.05407742038369179,
0.016885213553905487,
0.014831878244876862,
-0.07637543231248856,
0.... |
67 | Embodied Crowd Counting | [
"Runling Long",
"Yunlong Wang",
"Jia Wan",
"Xiang Deng",
"Xinting Zhu",
"Weili Guan",
"Antoni Chan",
"Liqiang Nie"
] | Occlusion is one of the fundamental challenges in crowd counting. In the community, various data-driven approaches have been developed to address this issue, yet their effectiveness is limited. This is mainly because most existing crowd counting datasets on which the methods are trained are based on passive cameras, restricting their ability to fully sense the environment.Recently, embodied navigation methods have shown significant potential in precise object detection in interactive scenes. These methods incorporate active camera settings, holding promise in addressing the fundamental issues in crowd counting. However, most existing methods are designed for indoor navigation, showing unknown performance in analyzing complex object distribution in large-scale scenes, such as crowds. Besides, most existing embodied navigation datasets are indoor scenes with limited scale and object quantity, preventing them from being introduced into dense crowd analysis. Based on this, a novel task, Embodied Crowd Counting (ECC), is proposed to count the number of persons in a large-scale scene actively. We then build up an interactive simulator, the Embodied Crowd Counting Dataset (ECCD), which enables large-scale scenes and large object quantities. A prior probability distribution approximating a realistic crowd distribution is introduced to generate crowds. Then, a zero-shot navigation method (ZECC) is proposed as a baseline. This method contains an MLLM-driven coarse-to-fine navigation mechanism, enabling active Z-axis exploration, and a normal-line-based crowd distribution analysis method for fine counting. Experimental results show that the proposed method achieves the best trade-off between counting accuracy and navigation cost. | poster | 2503.08367 | null | null | [] | [] | [] | [
0.00414729630574584,
-0.027716603130102158,
0.01586536504328251,
-0.006757046561688185,
0.02461111731827259,
0.015539009124040604,
0.01710420288145542,
0.023918692022562027,
-0.05646754056215286,
-0.0444561168551445,
-0.0292371679097414,
-0.038138821721076965,
-0.05409357324242592,
-0.0227... |
68 | Compress, Gather, and Recompute: REFORMing Long-Context Processing in Transformers | [
"Woomin Song",
"Sai Muralidhar Jayanthi",
"Srikanth Ronanki",
"Kanthashree Mysore Sathyendra",
"Jinwoo Shin",
"Aram Galstyan",
"Shubham Katiyar",
"Sravan Babu Bodapati"
] | As large language models increasingly gain popularity in real-world applications, processing extremely long contexts, often exceeding the model’s pre-trained context limits, has emerged as a critical challenge. While existing approaches to efficient long-context processing show promise, recurrent compression-based methods struggle with information preservation, whereas random access approaches require substantial memory resources. We introduce REFORM, a novel inference framework that efficiently handles long contexts through a two-phase approach. First, it incrementally processes input chunks while maintaining a compressed KV cache, constructs cross-layer context embeddings, and utilizes early exit strategy for improved efficiency. Second, it identifies and gathers essential tokens via similarity matching and selectively recomputes the KV cache. Compared to baselines, REFORM achieves over 50% and 27% performance gains on RULER and BABILong respectively at 1M context length. It also outperforms baselines on ∞-Bench, RepoEval, and MM-NIAH, demonstrating flexibility across diverse tasks and domains. Additionally, REFORM reduces inference time by 30% and peak memory usage by 5%, achieving both efficiency and superior performance. | poster | 2506.01215 | null | null | [] | [] | [] | [
-0.02584465965628624,
-0.04801769182085991,
-0.041235022246837616,
0.0027358834631741047,
0.04168868437409401,
0.044234104454517365,
-0.005667322315275669,
0.012368349358439445,
-0.023287972435355186,
-0.029234886169433594,
-0.036470357328653336,
0.019271232187747955,
-0.0414399653673172,
... |
69 | Curriculum Model Merging: Harmonizing Expert Chemical LLMs for Enhanced Cross-Task Generalization | [
"Baoyi He",
"Luotian Yuan",
"Ying Wei",
"Fei Wu"
] | The emergence of large language models (LLMs) prompts fine-tuning foundation LLMs to solve real-world chemical problems. However, these chemical LLMs are tailored to specific task formats or narrow content domains, which limits their capacity for comprehensive knowledge integration and cross-task generalization. Model merging has recently been proposed as a promising solution and demonstrated promising performance in the domain of natural language processing (NLP), enabling the integration of multiple expert LLMs into a unified model without requiring access to original training data or incurring significant computational costs. The widespread prevalence of in-house training data in the chemical domain underscores the need for practical and privacy-preserving model merging method for chemical LLM integration. % Although existing model merging methods %attempt to %minimize %harmonize interference among models and % have , However, two key characteristics of the chemical domain hinder effective model merging: 1) significant disparities among differentiated LLMs due to task-specific specializations, and 2) a highly imbalanced distribution of downstream functionality, with sparse model coverage for niche tasks and a surplus of models targeting widely studied problems. These factors intensify model inconsistencies, such as parameter interference and accumulated fine-tuning noise, ultimately degrade performance when integrating heterogeneous chemical LLMs. To this end, we propose Curriculum Model Merging (CMM), a method that constructs curriculum to progressively merge the expert chemical LLMs in a moderate and continual manner, aiming to harmonize the inconsistencies and mitigate the interference. Our comprehensive experiments on two benchmark datasets show that our proposed method concentrates task-specific expertise and outperforms the state-of-the-art methods by 29.03\% in terms of an overall average performance score. Moreover, CMM facilitates chemical knowledge generalization across prediction and generative tasks without sacrificing robustness, exhibiting promising merging performance under both expert-abundant and expert-sparse scenarios. | poster | null | null | null | [] | [] | [] | [
-0.013365387916564941,
-0.019961735233664513,
-0.024928901344537735,
0.0493999682366848,
0.06396392732858658,
-0.01689884252846241,
0.019987154752016068,
0.002038227394223213,
-0.011873426847159863,
-0.02405981533229351,
-0.026081670075654984,
0.031651757657527924,
-0.061614055186510086,
-... |
70 | Rebalancing Contrastive Alignment with Learnable Semantic Gaps in Text-Video Retrieval | [
"Jian Xiao",
"Zijie Song",
"Jialong Hu",
"Hao Cheng",
"Jia Li",
"Zhenzhen Hu",
"Richang Hong"
] | Recent advances in text-video retrieval have been largely driven by contrastive learning frameworks. However, existing methods overlook a key source of optimization tension: the separation between text and video distributions in the representation space—referred to as the modality gap—and the prevalence of false negatives in batch sampling. These factors lead to conflicting gradients under the InfoNCE loss, impeding stable alignment. To mitigate this, we propose GARE—a Gap-Aware Retrieval framework that introduces a learnable, pair-specific increment $\Delta_{ij}$ between text $t_i$ and video $v_j$ to offload the tension from the global anchor representation. We first derive the ideal form of $\Delta_{ij}$ via a coupling multivariate first-order Taylor approximation of the InfoNCE loss under a trust-region constraint, revealing it as a key mechanism for resolving gradient conflicts by guiding updates along a locally optimal descent direction in the coupled optimization landscape. Due to the expensive cost of directly approximate $\Delta_{ij}$, we introduce a lightweight neural module conditioned on the semantic gap between each video-text pair, enabling structure-aware correction guided by gradient supervision. To further stabilize learning and promote interpretability, we regularize $\Delta$ via three components: a trust-region constraint regularization to prevent oscillations, a directional diversity term to expand the semantic difference space, and an information bottleneck over $\Delta$ to restrict redundant information. Experiments across four retrieval benchmarks show that GARE consistently improves alignment accuracy and robustness to noisy supervision, confirming the effectiveness of gap-aware tension unloading. | poster | null | null | null | [] | [] | [] | [
0.016714055091142654,
-0.012733028270304203,
-0.012950748205184937,
0.08036302030086517,
0.021536992862820625,
-0.0009099894086830318,
0.04656541347503662,
-0.002893265336751938,
-0.02441183291375637,
-0.018476851284503937,
-0.007633114233613014,
0.03146977350115776,
-0.06346204876899719,
... |
71 | Non-convex entropic mean-field optimization via Best Response flow | [
"Razvan-Andrei Lascu",
"Mateusz Majka"
] | We study the problem of minimizing non-convex functionals on the space of probability measures, regularized by the relative entropy (KL divergence) with respect to a fixed reference measure, as well as the corresponding problem of solving entropy-regularized non-convex-non-concave min-max problems. We utilize the Best Response flow (also known in the literature as the fictitious play flow) and study how its convergence is influenced by the relation between the degree of non-convexity of the functional under consideration, the regularization parameter and the tail behaviour of the reference measure. In particular, we demonstrate how to choose the regularizer, given the non-convex functional, so that the Best Response operator becomes a contraction with respect to the $L^1$-Wasserstein distance, which then ensures the existence of its unique fixed point that is then showed to be the unique global minimizer for our optimization problem. This extends recent results where the Best Response flow was applied to solve convex optimization problems regularized by the relative entropy with respect to arbitrary reference measures, and with arbitrary values of the regularization parameter. Our results explain precisely how the assumption of convexity can be relaxed, at the expense of making a specific choice of the regularizer. Additionally, we demonstrate how these results can be applied in reinforcement learning in the context of policy optimization for bandit problems with softmax parametrized policies in the mean-field regime. | poster | 2505.22760 | null | null | [] | [] | [] | [
-0.0332493931055069,
-0.027188891544938087,
0.029551338404417038,
0.04296409338712692,
0.04979082569479942,
0.02583281882107258,
0.004629548639059067,
0.021831294521689415,
-0.025132808834314346,
-0.05960023030638695,
-0.018529947847127914,
0.012986650690436363,
-0.04789767786860466,
-0.00... |
72 | OLinear: A Linear Model for Time Series Forecasting in Orthogonally Transformed Domain | [
"Wenzhen Yue",
"Yong Liu",
"Haoxuan Li",
"Hao Wang",
"Xianghua Ying",
"Ruohao Guo",
"Bowei Xing",
"Ji Shi"
] | This paper presents $\mathbf{OLinear}$, a $\mathbf{linear}$-based multivariate time series forecasting model that operates in an $\mathbf{o}$rthogonally transformed domain. Recent forecasting models typically adopt the temporal forecast (TF) paradigm, which directly encode and decode time series in the time domain. However, the entangled step-wise dependencies in series data can hinder the performance of TF. To address this, some forecasters conduct encoding and decoding in the transformed domain using fixed, dataset-independent bases (e.g., sine and cosine signals in the Fourier transform). In contrast, we propose $\mathbf{OrthoTrans}$, a data-adaptive transformation based on an orthogonal matrix that diagonalizes the series' temporal Pearson correlation matrix. This approach enables more effective encoding and decoding in the decorrelated feature domain and can serve as a plug-in module to enhance existing forecasters. To enhance the representation learning for multivariate time series, we introduce a customized linear layer, $\mathbf{NormLin}$, which employs a normalized weight matrix to capture multivariate dependencies. Empirically, the NormLin module shows a surprising performance advantage over multi-head self-attention, while requiring nearly half the FLOPs. Extensive experiments on 24 benchmarks and 140 forecasting tasks demonstrate that OLinear consistently achieves state-of-the-art performance with high efficiency. Notably, as a plug-in replacement for self-attention, the NormLin module consistently enhances Transformer-based forecasters. The code and datasets are available at https://anonymous.4open.science/r/OLinear. | poster | 2505.08550 | https://github.com/jackyue1994/OLinear | null | [] | [] | [] | [
-0.017747769132256508,
-0.03783543407917023,
0.027367351576685905,
0.014735867269337177,
0.05368179455399513,
0.03788033127784729,
0.015162134543061256,
0.010035155341029167,
-0.019324062392115593,
-0.04433383420109749,
-0.012180443853139877,
-0.0012448306661099195,
-0.055709097534418106,
... |
73 | Noise Hypernetworks: Moving Test-Time Compute in Diffusion Models to Training | [
"Luca Eyring",
"Shyamgopal Karthik",
"Alexey Dosovitskiy",
"Nataniel Ruiz",
"Zeynep Akata"
] | The new paradigm of test-time scaling has yielded remarkable breakthroughs in Large Language Models (LLMs) (e.g.~reasoning models) and in generative vision models, allowing models to allocate additional computation during inference to effectively tackle increasingly complex problems. Despite the improvements of this approach, an important limitation emerges: the substantial increase in computation time makes the process slow and impractical for many applications. Given the success of this paradigm and its growing usage, we seek to preserve its benefits while eschewing the inference overhead. In this work we propose one solution to the critical problem of integrating test-time scaling knowledge into a model during post-training. Specifically, we replace reward guided test-time noise optimization in diffusion models with a Noise Hypernetwork that modulates initial input noise. We propose a theoretically grounded framework for learning this reward-tilted distribution for distilled generators, through a tractable noise-space objective that maintains fidelity to the base model while optimizing for desired characteristics. We show that our approach recovers a substantial portion of the quality gains from explicit test-time optimization at a fraction of the computational cost. | poster | null | null | null | [] | [] | [] | [
0.0038240968715399504,
-0.003000532742589712,
-0.006321100518107414,
0.06459168344736099,
0.046962983906269073,
0.04055578261613846,
0.035324208438396454,
-0.004947857465595007,
-0.008763374760746956,
-0.03237050399184227,
0.03100784868001938,
-0.0061977761797606945,
-0.03976071625947952,
... |
74 | Selective Learning for Deep Time Series Forecasting | [
"Yisong Fu",
"Zezhi Shao",
"Chengqing Yu",
"Yujie Li",
"Zhulin An",
"Qi Wang",
"Yongjun Xu",
"Fei Wang"
] | Benefiting from high capacity for capturing complex temporal patterns, deep learning (DL) has significantly advanced time series forecasting (TSF). However, deep models tend to suffer from severe overfitting due to the inherent vulnerability of time series to noise and anomalies. The prevailing DL paradigm uniformly optimizes all timesteps through the MSE loss and learns those uncertain and anomalous timesteps without difference, ultimately resulting in overfitting. To address this, we propose a novel selective learning strategy for deep TSF. Specifically, selective learning screens a subset of the whole timesteps to calculate the MSE loss in optimization, guiding the model to focus on generalizable timesteps while disregarding non-generalizable ones. Our framework introduces a dual-mask mechanism to target timesteps: (1) an uncertainty mask leveraging residual entropy to filter uncertain timesteps, and (2) an anomaly mask employing residual lower bound estimation to exclude anomalous timesteps. Extensive experiments across eight real-world datasets demonstrate that selective learning can significantly improve the predictive performance for typical state-of-the-art deep models, including 37.4% MSE reduction for Informer, 8.4% for TimesNet, and 6.5% for iTransformer. | poster | null | null | null | [] | [] | [] | [
-0.013199646025896072,
-0.023270947858691216,
0.0016114034224301577,
0.025527914986014366,
0.04869253560900688,
0.05370781570672989,
0.043502263724803925,
0.0020572294015437365,
-0.03249279037117958,
-0.046582795679569244,
0.0037303760182112455,
0.02528771385550499,
-0.06116538494825363,
0... |
75 | CyIN: Cyclic Informative Latent Space for Bridging Complete and Incomplete Multimodal Learning | [
"Ronghao Lin",
"Qiaolin He",
"Sijie Mai",
"Ying Zeng",
"Aolin Xiong",
"Li Huang",
"Yap-peng Tan",
"Haifeng Hu"
] | Multimodal machine learning, mimicking the human brain’s ability to integrate various modalities has seen rapid growth. Most previous multimodal models are trained on perfectly paired multimodal input to reach optimal performance. In real‑world deployments, however, the presence of modality is highly variable and unpredictable, causing the pre-trained models in suffering significant performance drops and fail to remain robust with dynamic missing modalities circumstances. In this paper, we present a novel Cyclic INformative Learning framework (CyIN) to bridge the gap between complete and incomplete multimodal learning. Specifically, we firstly builds an informative latent space by adopting token- and label-level Information Bottleneck (IB) cyclically among various modalities. Capturing task-related features with variational approximation, the informative bottleneck latents are purified for more efficient cross-modal interaction and multimodal fusion. Moreover, to supplement the missing information caused by incomplete multimodal input, we propose cross-modal cyclic translation by reconstruct the missing modalities with the remained ones through forward and reverse propagation process. With the help of the extracted and reconstructed informative latents, CyIN succeeds in jointly optimizing complete and incomplete multimodal learning in one unified model. Extensive experiments on 4 datasets demonstrate the superior performance of our method in both complete and diverse incomplete scenarios. | poster | null | null | null | [] | [] | [] | [
0.015692036598920822,
-0.02450178749859333,
-0.013273391872644424,
0.058086082339286804,
0.04695693403482437,
0.008746236562728882,
0.01173874456435442,
0.03005606308579445,
-0.04639524966478348,
-0.01662602834403515,
-0.03288876637816429,
0.01129361055791378,
-0.056958816945552826,
0.0118... |
76 | Interaction-Centric Knowledge Infusion and Transfer for Open Vocabulary Scene Graph Generation | [
"Lin Li",
"Chuhan ZHANG",
"Dong Zhang",
"Chong Sun",
"Chen Li",
"Long Chen"
] | Open-vocabulary scene graph generation (OVSGG) extends traditional SGG by recognizing novel objects and relationships beyond predefined categories, leveraging the knowledge from pre-trained large-scale models. Existing OVSGG methods always adopt a two-stage pipeline: 1) Infusing knowledge into large-scale models via pre-training on large datasets; 2) Transferring knowledge from pre-trained models with fully annotated scene graphs during supervised fine-tuning. However, due to the lack of explicit interaction modeling, these methods struggle to distinguish between interacting and non-interacting instances of the same object category, which significantly exacerbates relation pair mismatches. To this end, in this paper, we propose an interACtion-Centric end-to-end OVSGG framework (ACC) in an interaction-driven paradigm to minimize these mismatches. For interaction-centric knowledge infusion, ACC employs a bidirectional interaction prompt for robust pseudo-supervision generation to enhance the model's interaction knowledge. For interaction-centric knowledge transfer, ACC first adopts interaction-guided query selection that prioritizes pairing interacting objects to reduce interference from non-interacting ones. Then, it integrates interaction-consistent knowledge distillation to bolster robustness by pushing relational foreground away from the background while retaining general knowledge. Extensive experimental results on three benchmarks show that ACC achieves state-of-the-art performance, demonstrating the potential of interaction-centric paradigms for real-world applications. | poster | null | null | null | [] | [] | [] | [
-0.023697935044765472,
-0.009951629675924778,
0.05005264654755592,
0.020781463012099266,
0.034026969224214554,
-0.008291463367640972,
0.04905814304947853,
0.008704342879354954,
0.014272869564592838,
-0.016466211527585983,
-0.03090555965900421,
0.028392137959599495,
-0.08035819232463837,
0.... |
77 | On the Integration of Spatial-Temporal Knowledge: A Lightweight Approach to Atmospheric Time Series Forecasting | [
"Yisong Fu",
"Fei Wang",
"Zezhi Shao",
"Boyu Diao",
"Lin Wu",
"Zhulin An",
"Chengqing Yu",
"Yujie Li",
"Yongjun Xu"
] | Transformers have gained attention in atmospheric time series forecasting (ATSF) for their ability to capture global spatial-temporal correlations. However, their complex architectures lead to excessive parameter counts and extended training times, limiting their scalability to large-scale forecasting. In this paper, we revisit ATSF from a theoretical perspective of atmospheric dynamics and uncover a key insight: spatial-temporal position embedding (STPE) can inherently model spatial-temporal correlations even without attention mechanisms. Its effectiveness arises from integrating geographical coordinates and temporal features, which are intrinsically linked to atmospheric dynamics. Based on this, we propose **STELLA**, a **S**patial-**T**emporal knowledge **E**mbedded **L**ightweight mode**L** for ASTF, utilizing only STPE and an MLP architecture in place of Transformer layers. With 10k parameters and one hour of training, STELLA achieves superior performance on five datasets compared to other advanced methods. The paper emphasizes the effectiveness of spatial-temporal knowledge integration over complex architectures, providing novel insights for ATSF. | poster | 2408.09695 | null | null | [] | [] | [] | [
0.02351103164255619,
-0.03387979418039322,
0.02483379654586315,
0.021668659523129463,
0.025519073009490967,
0.03659374266862869,
0.041329171508550644,
0.00021078753343317658,
-0.03005489893257618,
-0.0443604551255703,
-0.020024850964546204,
0.008486934006214142,
-0.05712788924574852,
0.004... |
78 | Path-specific effects for pulse-oximetry guided decisions in critical care | [
"Kevin Zhang",
"Yonghan Jung",
"Divyat Mahajan",
"Karthikeyan Shanmugam",
"Shalmali Joshi"
] | Identifying and measuring biases associated with sensitive attributes is a crucial consideration in healthcare to prevent treatment disparities. One prominent issue is inaccurate pulse oximeter readings, which tend to overestimate oxygen saturation for dark-skinned patients and misrepresent supplemental oxygen needs. Most existing research has revealed *statistical disparities* linking device errors to patient outcomes in Intensive Care Units (ICUs) without causal formalization. In contrast, this study *causally* investigates how racial discrepancies in oximetry measurements affect invasive ventilation in intensive care units (ICUs). We introduce a causal inference-based approach using *path-specific effects* to isolate the impact of bias by race on clinical decision-making. To estimate these effects, we leverage a doubly robust estimator, propose its self-normalized variant for improved sample efficiency, and provide novel finite-sample guarantees. Our methodology is validated on semi-synthetic data and applied to two large real-world health datasets: MIMIC-IV and eICU. Contrary to prior work, our analysis reveals minimal impact of racial discrepancies on invasive ventilation rates. However, path-specific effects mediated by oxygen saturation disparity were more pronounced on ventilation duration and differed by dataset. Our work provides a novel and practical pipeline for investigating potential disparities in the ICU and, more crucially, highlights the necessity of causal methods to robustly assess fairness in decision-making. | poster | 2506.12371 | null | null | [] | [] | [] | [
0.008111752569675446,
-0.00938156247138977,
-0.05360705032944679,
0.0029024085961282253,
0.05745077505707741,
0.045557063072919846,
0.06835025548934937,
0.04384239390492439,
0.004212590865790844,
-0.05759977549314499,
0.016937658190727234,
-0.014275893568992615,
-0.08000466227531433,
-0.02... |
79 | InfMasking: Unleashing Synergistic Information \\ by Contrastive Multimodal Interactions | [
"Liangjian Wen",
"Qun Dai",
"Yong Dai",
"Jianzhuang Liu",
"Jiangtao Zheng",
"Dongkai Wang",
"Zhao Kang",
"Jun Wang",
"Zenglin Xu",
"Jiang Duan"
] | In multimodal representation learning, synergistic interactions between modalities not only provide complementary information but also create unique outcomes through specific interaction patterns that no single modality could achieve alone. Existing methods may struggle to effectively capture the full spectrum of synergistic information, leading to suboptimal performance in tasks where such interactions are critical. This is particularly problematic because synergistic information constitutes the fundamental value proposition of multimodal representation. To address this challenge, we introduce InfMasking, a contrastive synergistic information extraction method designed to enhance synergistic information through an Infinite Masking strategy. InfMasking stochastically occludes most features from each modality during fusion, preserving only partial information to create representations with varied synergistic patterns. Unmasked fused representations are then aligned with masked ones through mutual information maximization to encode comprehensive synergistic information. This infinite masking strategy enables capturing richer interactions by exposing the model to diverse partial modality combinations during training. As computing mutual information estimates with infinite masking is computationally prohibitive, we derive an InfMasking loss to approximate this calculation. Through controlled experiments, we demonstrate that InfMasking effectively enhances synergistic information between modalities. In evaluations on large-scale real-world datasets, InfMasking achieves state-of-the-art performance across seven benchmarks. | spotlight | 2509.25270 | https://github.com/brightest66/InfMasking | null | [] | [] | [] | [
0.0025340982247143984,
-0.003961007576435804,
-0.007449930999428034,
0.034645937383174896,
0.04088214412331581,
-0.0008044203859753907,
0.023103052750229836,
0.00767563795670867,
-0.06622260808944702,
-0.019195402041077614,
-0.0073036556132137775,
0.012889462523162365,
-0.07698565721511841,
... |
80 | UniGen: Enhanced Training & Test-Time Strategies for Unified Multimodal Understanding and Generation | [
"Rui Tian",
"Mingfei Gao",
"Mingze Xu",
"Jiaming Hu",
"Jiasen Lu",
"Zuxuan Wu",
"Yinfei Yang",
"Afshin Dehghan"
] | We introduce UniGen, a unified multimodal large language model (MLLM) capable of image understanding and generation. We study the full training pipeline of UniGen from a data-centric perspective, including multi-stage pre-training, supervised fine-tuning, and direct preference optimization. More importantly, we propose a new Chain-of-Thought Verification (CoT-V) strategy for test-time scaling, which significantly boosts UniGen's image generation quality using a simple Best-of-N test-time strategy. Specifically, CoT-V enables UniGen to act as both image generator and verifier at test time, assessing the semantic alignment between a text prompt and its generated image in a step-by-step CoT manner. Trained entirely on open-source datasets across all stages, UniGen achieves state-of-the-art performance on a range of image understanding and generation benchmarks, with a final score of 0.78 on GenEval and 85.19 on DPG-Bench. Through extensive ablation studies, our work provides actionable insights and addresses key challenges in the full life cycle of building unified MLLMs, contributing meaningful directions to the future research. | poster | 2505.14682 | null | null | [] | [] | [] | [
0.01636873371899128,
-0.031271323561668396,
0.021039122715592384,
0.05573428422212601,
0.03212164714932442,
0.013030659407377243,
0.021963104605674744,
0.009948237799108028,
-0.006730770692229271,
-0.011750472709536552,
0.00109754444565624,
0.01813638210296631,
-0.07438857108354568,
0.0086... |
81 | Regret Analysis of Average-Reward Unichain MDPs via an Actor-Critic Approach | [
"Swetha Ganesh",
"Vaneet Aggarwal"
] | Actor-Critic methods are widely used for their scalability, yet existing theoretical guarantees for infinite-horizon average-reward Markov Decision Processes (MDPs) often rely on restrictive ergodicity assumptions. We propose NAC-B, a Natural Actor-Critic with Batching, that achieves order-optimal regret of \$\tilde{O}(\sqrt{T})\$ in infinite-horizon average-reward MDPs under the unichain assumption, which permits both transient states and periodicity. This assumption is among the weakest under which the classic policy gradient theorem remains valid for average-reward settings. NAC-B employs function approximation for both the actor and the critic, enabling scalability to problems with large state and action spaces. The use of batching in our algorithm helps mitigate potential periodicity in the MDP and reduces stochasticity in gradient estimates, and our analysis formalizes these benefits through the introduction of the constants $C_{\text{hit}}$ and $C_{\text{tar}}$, which characterize the rate at which empirical averages over Markovian samples converge to the stationary distribution. | poster | 2505.19986 | null | null | [] | [] | [] | [
-0.06101256608963013,
-0.041662417352199554,
-0.029190024361014366,
0.044190067797899246,
0.046417053788900375,
0.02210015058517456,
0.01836724951863289,
0.006231789011508226,
-0.033105697482824326,
-0.05615061894059181,
-0.005755631718784571,
0.004362663719803095,
-0.07786750048398972,
-0... |
82 | TreeSplat: Mergeable Tree for Deformable Gaussian Splatting | [
"Qiuhong Shen",
"Xingyi Yang",
"Xinchao Wang"
] | Dynamic 3D scene reconstruction from multi-view videos demands representation to model complex deformations at scale. Current Gaussian Splatting based methods often either suffer from significant computation cost due to dense MLP-based modeling or explicit modeling deformation of each Gaussian independently. However, the dynamics of objects within a scene are typically hierarchical and exhibit structural correlations. To leverage these structural priors into the representation, we introduce **TreeSplat**, a **Tree** data structure for deformable Gaussian **Splat**ting. In TreeSplat, as the name suggests, motions of Gaussian are represented hierarchically within a tree. Each node learns coefficients for time-varying basis functions, defining a part of the motion. The full motion for any given Gaussian is then determined by accumulating these transformations along the tree path from its leaf node to the root node. This tree isn't predefined; instead, it is constructed adaptively alongside Gaussian densification, where cloning or splitting a Gaussian correspondingly creates new leaf nodes. One central property of TreeSplat is its mergeability; after optimization during training, the hierarchical motion parameters for each Gaussian can be efficiently consolidated. By performing this merging step before test time, we eliminate the need to traverse the tree explicitly for each Gaussian during rendering. This results in dramatically faster rendering over 200 FPS and compact storage, while maintaining state-of-the-art rendering quality. Experiments on diverse synthetic and real-world datasets validate these advantages. | poster | null | null | null | [] | [] | [] | [
0.0015558501472696662,
-0.03231716528534889,
0.013307534158229828,
0.04777367785573006,
0.009764793328940868,
0.03987479582428932,
0.022225502878427505,
0.014180563390254974,
-0.022509761154651642,
-0.05443588271737099,
-0.030348965898156166,
-0.038762371987104416,
-0.05093426629900932,
-0... |
83 | 1000+ FPS 4D Gaussian Splatting for Dynamic Scene Rendering | [
"Yuheng Yuan",
"Qiuhong Shen",
"Xingyi Yang",
"Xinchao Wang"
] | 4D Gaussian Splatting (4DGS) has recently gained considerable attention as a method for reconstructing dynamic scenes. Despite achieving superior quality, 4DGS typically requires substantial storage and suffers from slow rendering speed. In this work, we delve into these issues and identify two key sources of temporal redundancy. (Q1) \textbf{Short-Lifespan Gaussians}: 4DGS uses a large portion of Gaussians with short temporal span to represent scene dynamics, leading to an excessive number of Gaussians. (Q2) \textbf{Inactive Gaussians}: When rendering, only a small subset of Gaussians contributes to each frame. Despite this, all Gaussians are processed during rasterization, resulting in redundant computation overhead. To address these redundancies, we present \textbf{4DGS-1K}, which runs at over 1000 FPS on modern GPUs. For Q1, we introduce the Spatial-Temporal Variation Score, a new pruning criterion that effectively removes short-lifespan Gaussians while encouraging 4DGS to capture scene dynamics using Gaussians with longer temporal spans. For Q2, we store a mask for active Gaussians across consecutive frames, significantly reducing redundant computations in rendering. Compared to vanilla 4DGS, our method achieves a $41\times$ reduction in storage and $9\times$ faster rasterization speed on complex dynamic scenes, while maintaining comparable visual quality. | poster | 2503.16422 | null | https://4dgs-1k.github.io/ | [] | [] | [] | [
-0.0025351077783852816,
-0.006246836856007576,
0.043932560831308365,
0.058671798557043076,
0.005747961811721325,
0.03780023753643036,
0.01738830842077732,
0.020202437415719032,
-0.04046609625220299,
-0.05361613258719444,
-0.0076958173885941505,
-0.02667398378252983,
-0.059488777071237564,
... |
84 | Test3R: Learning to Reconstruct 3D at Test Time | [
"Yuheng Yuan",
"Qiuhong Shen",
"Shizun Wang",
"Xingyi Yang",
"Xinchao Wang"
] | Dense matching methods like DUST3R regress pairwise pointmaps for 3D reconstruction. However, the reliance on pairwise prediction and the limited generalization capability inherently restrict the global geometric consistency. In this work, we introduce \textbf{Test3R}, a surprisingly simple test-time learning technique that significantly boosts geometric accuracy. Using image triplets ($I_1,I_2,I_3$), Test3R generates reconstructions from pairs ($I_1,I_2$) and ($I_1,I_3$). The core idea is to optimize the network at test time via a self-supervised objective: maximizing the geometric consistency between these two reconstructions relative to the common image $I_1$. This ensures the model produces cross-pair consistent outputs, regardless of the inputs. Extensive experiments demonstrate that our technique significantly outperforms previous state-of-the-art methods on the 3D reconstruction and multi-view depth estimation tasks. Moreover, it is universally applicable and nearly cost-free, making it easily applied to other models and implemented with minimal test-time training overhead and parameter footprint. | poster | 2506.13750 | https://github.com/nopQAQ/Test3R | null | [] | [] | [] | [
0.03479780629277229,
0.011392438784241676,
0.0030747309792786837,
0.03272772952914238,
0.024794666096568108,
0.043982334434986115,
0.015806809067726135,
0.013638238422572613,
-0.015886854380369186,
-0.06647180765867233,
0.00795132014900446,
-0.0002748797705862671,
-0.04976772144436836,
0.0... |
85 | Exploring Neural Granger Causality with xLSTMs: Unveiling Temporal Dependencies in Complex Data | [
"Harsh Poonia",
"Felix Divo",
"Kristian Kersting",
"Devendra Singh Dhami"
] | Causality in time series can be difficult to determine, especially in the presence of non-linear dependencies. The concept of Granger causality helps analyze potential relationships between variables, thereby offering a method to determine whether one time series can predict—Granger cause—future values of another. Although successful, Granger causal methods still struggle with capturing long-range relations between variables. To this end, we leverage the recently successful Extended Long Short-Term Memory (xLSTM) architecture and propose Granger causal xLSTMs (GC-xLSTM). It first enforces sparsity between the time series components by using a novel dynamic loss penalty on the initial projection. Specifically, we adaptively improve the model and identify sparsity candidates. Our joint optimization procedure then ensures that the Granger causal relations are recovered robustly. Our experimental evaluation on six diverse datasets demonstrates the overall efficacy of our proposed GC-xLSTM model. | poster | 2502.09981 | null | null | [] | [] | [] | [
-0.010957252234220505,
-0.024442534893751144,
-0.010079051367938519,
0.030630286782979965,
0.04653097316622734,
0.04125400632619858,
0.024531498551368713,
0.009364068508148193,
-0.03540399670600891,
-0.030892184004187584,
0.009803054854273796,
0.0000717660368536599,
-0.05263455584645271,
0... |
86 | Angles Don’t Lie: Unlocking Training‑Efficient RL Through the Model’s Own Signals | [
"Qinsi Wang",
"Jinghan Ke",
"Hancheng Ye",
"Yueqian Lin",
"Yuzhe Fu",
"Jianyi Zhang",
"Kurt Keutzer",
"Chenfeng Xu",
"Yiran Chen"
] | Current Reinforcement Fine-tuning (RFT) paradigms for Large Language Models (LLMs) suffer from sample inefficiency due to the redundant exposure of identical queries under uniform data sampling. While previous work has explored curriculum learning via heuristic difficulty metrics, these strategies exhibit limitations by neglecting the intrinsic learning signals generated by the model itself, thus leading to suboptimal training regimes. In this paper, we identify a model-inherent signal termed *angle concentration* that effectively reflects an LLM's capacity to learn from specific data. We theoretically and empirically demonstrate a correlation between the angular distribution of token hidden state vectors and the resulting gradient, revealing a learning preference for data exhibiting higher angle concentration. Inspired by this finding, we propose GAIN-RL, a Gradient-driven Angle-Informed Navigated RL framework. By leveraging the model's intrinsic angle concentration signal, GAIN-RL dynamically selects training data in each epoch, ensuring consistently impactful gradient updates and thus significantly enhancing overall training efficiency. Empirical evaluations show that GAIN-RL (GRPO) achieves over a 2.5$\times$ acceleration in training efficiency across diverse mathematical and coding tasks and varying model scales. Furthermore, GAIN-RL (GRPO)'s efficient sampling yields data-efficient training, achieving better performance with half the original data compared to vanilla GRPO with full training data. | spotlight | null | null | null | [] | [] | [] | [
-0.028968345373868942,
-0.013903058134019375,
0.018788985908031464,
0.022557387128472328,
0.022066816687583923,
0.02894545905292034,
0.04059319198131561,
0.00503338361158967,
-0.0511966235935688,
-0.00035111833130940795,
-0.003574170172214508,
0.03836223483085632,
-0.052435606718063354,
-0... |
87 | Image Editing As Programs with Diffusion Models | [
"Yujia Hu",
"Songhua Liu",
"Zhenxiong Tan",
"Xingyi Yang",
"Xinchao Wang"
] | While diffusion models have achieved remarkable success in text-to-image generation, they encounter significant challenges with instruction-driven image editing. Our research highlights a key challenge: these models particularly struggle with structurally-inconsistent edits that involve substantial layout changes. To address this gap, we introduce Image Editing As Programs (IEAP), a unified image editing framework built upon the Diffusion Transformer (DiT) architecture. Specifically, IEAP deals with complex instructions by decomposing them into a sequence of programmable atomic operations. Each atomic operation manages a specific type of structurally consistent edit; when sequentially combined, IEAP enables the execution of arbitrary, structurally-inconsistent transformations. This reductionist approach enables IEAP to robustly handle a wide spectrum of edits, encompassing both structurally-consistent and -inconsistent changes. Extensive experiments demonstrate that IEAP significantly outperforms state-of-the-art methods on standard benchmarks across various editing scenarios. In these evaluations, our framework delivers superior accuracy and semantic fidelity, particularly for complex, multi-step instructions. | poster | 2506.04158 | https://github.com/YujiaHu1109/IEAP | https://yujiahu1109.github.io/IEAP/ | [
"Cicici1109/IEAP"
] | [
"Cicici1109/IEAP"
] | [] | [
0.021243654191493988,
-0.003844137769192457,
-0.03464822098612785,
0.061760105192661285,
0.0520746223628521,
0.030880965292453766,
0.020028192549943924,
0.008630071766674519,
-0.0034783382434397936,
-0.05769478529691696,
-0.009849436581134796,
-0.015177702531218529,
-0.05140269175171852,
-... |
88 | Aura Attention: $\mathcal O(n \log n)$ Sparse Attention for Long Video Generation | [
"XINGYANG LI",
"Muyang Li",
"Tianle Cai",
"Haocheng Xi",
"Shuo Yang",
"Yujun Lin",
"Lvmin Zhang",
"Songlin Yang",
"Jinbo Hu",
"Kelly Peng",
"Maneesh Agrawala",
"Ion Stoica",
"Kurt Keutzer",
"Song Han"
] | Recent advances in diffusion models have enabled high-quality video generation, but the additional temporal dimension significantly increases computational costs, making training and inference on long videos prohibitively expensive. In this paper, we identify a phenomenon we call \textit{Spatiotemporal Energy Decay} in video diffusion models: post-softmax attention scores decrease as spatial and temporal distance between tokens increase, akin to the physical decay of signal or waves over space and time in nature. Motivated by this, we propose \textit{\method}, a scalable sparse attention mechanism with $\mathcal{O}(n \log n)$ complexity that translates energy decay into exponentially decaying compute density. \method employs a simple, static attention mask where each token attends to spatially nearby tokens, with the attention window size shrinking with temporal distance. Moreover, \method allows pre-trained video diffusion models to extend their generation length with minimal fine-tuning with LoRA. Extensive experiments show that \method maintains video quality across Wan2.1-14B, HunyuanVideo, achieving up to a 1.9× speedup over full attention. With minimal fine-tuning, it enables video generation up to 4× longer while reducing training costs by up to 4.4× compared to direct fine-tuning and accelerating inference by 3.7× compared to full attention inference. Code and models will be released upon publication. | poster | null | null | null | [] | [] | [] | [
0.034862618893384933,
-0.00773136829957366,
0.02601141668856144,
0.029255777597427368,
0.02908935211598873,
0.02808031067252159,
0.02555537410080433,
0.012401510030031204,
-0.031378086656332016,
-0.04454432427883148,
0.004092770628631115,
-0.019305331632494926,
-0.039955172687768936,
0.027... |
89 | Influence Functions for Edge Edits in Non-Convex Graph Neural Networks | [
"Jaeseung Heo",
"Kyeongheung Yun",
"Seokwon Yoon",
"MoonJeong Park",
"Jungseul Ok",
"Dongwoo Kim"
] | Understanding how individual edges influence the behavior of graph neural networks (GNNs) is essential for improving their interpretability and robustness. Graph influence functions have emerged as promising tools to efficiently estimate the effects of edge deletions without retraining. However, existing influence prediction methods rely on strict convexity assumptions, exclusively consider the influence of edge deletions while disregarding edge insertions, and fail to capture changes in message propagation caused by these modifications. In this work, we propose a proximal Bregman response function specifically tailored for GNNs, relaxing the convexity requirement and enabling accurate influence prediction for standard neural network architectures. Furthermore, our method explicitly accounts for message propagation effects and extends influence prediction to both edge deletions and insertions in a principled way. Experiments with real-world datasets demonstrate accurate influence predictions for different characteristics of GNNs. We further demonstrate that the influence function is versatile in applications such as graph rewiring and adversarial attacks. | poster | 2506.04694 | null | null | [] | [] | [] | [
-0.0195511095225811,
-0.02566024847328663,
-0.0011339356424286962,
0.06067647039890289,
0.036580175161361694,
0.0292117428034544,
0.017486566677689552,
0.010225608944892883,
-0.008751687593758106,
-0.04128332808613777,
0.012292266823351383,
0.010516499169170856,
-0.07535093277692795,
-0.00... |
90 | Elastic ViTs from Pretrained Models without Retraining | [
"Walter Simoncini",
"Michael Dorkenwald",
"Tijmen Blankevoort",
"Cees Snoek",
"Yuki Asano"
] | Vision foundation models achieve remarkable performance but are only available in a limited set of pre-determined sizes, forcing sub-optimal deployment choices under real-world constraints. We introduce a novel post-pretraining structured pruning method that enables elastic inference across a continuum of compute budgets. Our approach combines gradient information with cross-network structure correlations, efficiently approximated through an evolutionary algorithm, does not require labeled data, generalizes to models without a classification head, and is retraining free. Experiments on DINO and AugReg models demonstrate superior performance over state of the art methods across various sparsities, requiring less than five minutes on a A100 GPU to generate elastic models that can be adjusted to any computational budget. Our key contributions include an efficient pruning strategy for pretrained Vision Transformers, a novel evolutionary approximation of Hessian off-diagonal structures, and a self-supervised importance scoring mechanism that maintains strong performance without requiring retraining nor labels. | poster | null | null | null | [] | [] | [] | [
0.007354293949902058,
-0.030562227591872215,
0.025025155395269394,
0.03104228340089321,
0.028825145214796066,
0.060181114822626114,
0.0070723434910178185,
0.03006516769528389,
-0.02622370608150959,
-0.07283040881156921,
0.009746128693223,
0.03060808964073658,
-0.07255937904119492,
0.003821... |
91 | SGAR: Structural Generative Augmentation for 3D Human Motion Retrieval | [
"Jiahang Zhang",
"Lilang Lin",
"Shuai Yang",
"Jiaying Liu"
] | 3D human motion-text retrieval is essential for accurate motion understanding, targeted at cross-modal alignment learning. Existing methods typically align the global motion-text concepts directly, suffering from sub-optimal generalization due to the uncertainty of correspondence learning between multiple motion concepts coupled in a single motion/text sequence. Therefore, we study the explicit fine-grained concept decomposition for alignment learning and present a novel framework, Structural Generative Augmentation for 3D Human Motion Retrieval (SGAR), to enable generation-augmented retrieval. Specifically, relying on the strong priors of existing large language model (LLM) assets, we effectively decompose human motions structurally into subtler semantic units, \ie, body parts, for fine-grained motion modeling. Based on this, we develop part-mixture learning to better decouple the local motion concept learning, boosting part-level alignment. Moreover, a directional relation alignment strategy exploiting the correspondence between full-body and part motions is incorporated to regularize feature manifold for better consistency. Extensive experiments on three benchmarks, including motion-text retrieval as well as recognition and generation applications, demonstrate the superior performance and promising transferability of our method. | poster | null | null | null | [] | [] | [] | [
0.016272295266389847,
-0.01942581869661808,
-0.017023345455527306,
0.0555490180850029,
0.01869324781000614,
0.011360653676092625,
0.040791384875774384,
0.025675518438220024,
-0.025999046862125397,
-0.02863915078341961,
-0.02220752276480198,
-0.017794478684663773,
-0.08146698772907257,
-0.0... |
92 | T-REGS: Minimum Spanning Tree Regularization for Self-Supervised Learning | [
"Julie Mordacq",
"David Loiseaux",
"Vicky Kalogeiton",
"Steve OUDOT"
] | Self-supervised learning (SSL) has emerged as a powerful paradigm for learning representations without labeled data, often by enforcing invariance to input transformations such as rotations or blurring.Recent studies have highlighted two pivotal properties for effective representations: (i) avoiding dimensional collapse-where the learned features occupy only a low-dimensional subspace, and (ii) enhancing uniformity of the induced distribution.In this work, we introduce T-REGS, a simple regularization framework for SSL based on the length of the Minimum Spanning Tree (MST) over the learned representation.We provide theoretical analysis demonstrating that T-REGS simultaneously mitigates dimensional collapse and promotes distribution uniformity on arbitrary compact Riemannian manifolds.Several experiments on synthetic data and on classical SSL benchmarks validate the effectiveness of our approach at enhancing representation quality. | spotlight | null | null | null | [] | [] | [] | [
0.03205886855721474,
-0.028401821851730347,
0.006817216984927654,
0.0403568334877491,
0.02919861115515232,
0.030436921864748,
0.046736571937799454,
-0.015000435523688793,
-0.04732092097401619,
-0.029049230739474297,
-0.005479604471474886,
-0.0120310690253973,
-0.06139228120446205,
0.027444... |
93 | Noise Matters: Optimizing Matching Noise for Diffusion Classifiers | [
"Yanghao Wang",
"Long Chen"
] | Although today's pretrained discriminative vision-language models (e.g., CLIP) have demonstrated strong perception abilities, such as zero-shot image classification, they also suffer from the bag-of-words problem and spurious bias. To mitigate these problems, some pioneering studies leverage powerful generative models (e.g., pretrained diffusion models) to realize generalizable image classification, dubbed Diffusion Classifier (DC). Specifically, by randomly sampling a Gaussian noise, DC utilizes the differences of denoising effects with different category conditions to classify categories. Unfortunately, an inherent and notorious weakness of existing DCs is noise instability: different random sampled noises lead to significant performance changes. To achieve stable classification performance, existing DCs always ensemble the results of hundreds of sampled noises, which significantly reduces the classification speed. To this end, we firstly explore the role of noise in DC, and conclude that: there are some ``good noises'' that can relieve the instability. Meanwhile, we argue that these good noises should meet two principles: 1) Frequency Matching: noise should destroy the specific frequency signals; 2) Spatial Matching: noise should destroy the specific spatial areas. Regarding both principles, we propose a novel Noise Optimization method to learn matching (i.e., good) noise for DCs: NoOp. For frequency matching, NoOp first optimizes a dataset-specific noise: Given a dataset and a timestep $t$, optimize one randomly initialized parameterized noise. For Spatial Matching, NoOp trains a Meta-Network that adopts an image as input and outputs image-specific noise offset. The sum of optimized noise and noise offset will be used in DC to replace random noise. Extensive ablations on various datasets demonstrated the effectiveness of NoOp. It is worth noting that our noise optimization is orthogonal to existing optimization methods (e.g., prompt tuning), our NoOP can even benefit from these methods to further boost performance. | poster | 2508.11330 | null | null | [] | [] | [] | [
-0.0100114019587636,
0.008453630842268467,
-0.005019061733037233,
0.059122826904058456,
0.03302552178502083,
0.06295917928218842,
0.010932715609669685,
-0.005247118417173624,
-0.018209092319011688,
-0.07385027408599854,
-0.029143115505576134,
0.01051815040409565,
-0.08021977543830872,
0.01... |
94 | Quantifying the Potential of Control Algorithms through Large Language Models | [
"Lianchen Jia",
"Chaoyang Li",
"Qian Houde",
"Tianchi Huang",
"Jiangchuan Liu",
"Lifeng Sun"
] | Control algorithms in production environments typically require domain experts to tune them for specific scenarios. However, existing research predominantly focuses on algorithmic performance under ideal or default configurations, overlooking the critical aspect of tuning potential. To bridge this gap, we introduce \texttt{Crucible}, a novel framework leveraging large language models to quantitatively evaluate the tuning potential of control algorithms. We demonstrate \texttt{Crucible}'s effectiveness through case studies of adaptive bitrate algorithms from networking and scheduling algorithms from systems domains. Our experimental results reveal that \texttt{Crucible} not only quantifies the tunable space across different algorithms but also enables the optimization of existing algorithms by using potential as a new optimization direction, ultimately enhancing performance outcomes. Our code is available in the supplementary materials. | poster | null | null | null | [] | [] | [] | [
-0.04712753742933273,
-0.00997680053114891,
-0.020388737320899963,
0.05029460787773132,
0.042053528130054474,
0.007767238188534975,
0.03444763645529747,
0.03344470262527466,
-0.016583330929279327,
-0.026646560057997704,
-0.019811579957604408,
0.004326962865889072,
-0.08064856380224228,
-0.... |
95 | Smooth Sailing: Lipschitz-Driven Uncertainty Quantification for Spatial Associations | [
"David Burt",
"Renato Berlinghieri",
"Stephen Bates",
"Tamara Broderick"
] | Estimating associations between spatial covariates and responses — rather than merely predicting responses — is central to environmental science, epidemiology, and economics. For instance, public health officials might be interested in whether air pollution has a strictly positive association with a health outcome, and the magnitude of any effect. Standard machine learning methods often provide accurate predictions but offer limited insight into covariate-response relationships. And we show that existing methods for constructing confidence (or credible) intervals for associations fail to provide nominal coverage in the face of model misspecification and distribution shift — despite both being essentially always present in spatial problems. We introduce a method that constructs valid frequentist confidence intervals for associations in spatial settings. Our method requires minimal assumptions beyond a form of spatial smoothness. In particular, we do not require model correctness or covariate overlap between training and target locations. Our approach is the first to guarantee nominal coverage in this setting and outperforms existing techniques in both real and simulated experiments. | poster | 2502.06067 | null | null | [] | [] | [] | [
0.019440293312072754,
0.006455260328948498,
-0.031933244317770004,
0.03961295634508133,
0.03632363677024841,
0.04356081411242485,
0.053068678826093674,
0.011160763911902905,
-0.020006468519568443,
-0.051678963005542755,
0.00055441859876737,
-0.01856706105172634,
-0.0854848101735115,
-0.009... |
96 | The quest for the GRAph Level autoEncoder (GRALE) | [
"Paul Krzakala",
"Gabriel Melo",
"Charlotte Laclau",
"Florence d'Alché-Buc",
"Rémi Flamary"
] | Although graph-based learning has attracted a lot of attention, graph representation learning is still a challenging task whose resolution may impact key application fields such as chemistry or biology. To this end, we introduce GRALE, a novel graph autoencoder that encodes and decodes graphs of varying sizes into a shared embedding space. GRALE is trained using an Optimal Transport-inspired loss that compares the source and reconstructed graphs and leverages a differentiable matching module, which is trained jointly with the encoder and decoder. The proposed attention-based architecture relies on Evoformer, the core component of AlphaFold, which we extend to support both graph encoding and decoding. We show, in numerical experiments on simulated and molecular data, that GRALE enables a highly general form of pre-training, applicable to a wide range of downstream tasks, from classification and regression to more complex tasks such as graph interpolation, editing, matching, and prediction. | poster | 2505.22109 | null | null | [] | [] | [] | [
-0.010212933644652367,
-0.002951559145003557,
-0.02175552025437355,
0.06569128483533859,
0.025702355429530144,
0.02810877561569214,
0.03930370509624481,
0.010946448892354965,
0.011440993286669254,
-0.059562429785728455,
0.03655679523944855,
-0.027456054463982582,
-0.07656705379486084,
0.00... |
97 | Word-Level Emotional Expression Control in Zero-Shot Text-to-Speech Synthesis | [
"Tianrui Wang",
"Haoyu Wang",
"Meng Ge",
"Cheng Gong",
"Chunyu Qiang",
"Ziyang Ma",
"Zikang Huang",
"Guanrou Yang",
"Xiaobao Wang",
"Eng-Siong Chng",
"Xie Chen",
"Longbiao Wang",
"Jianwu Dang"
] | While emotional text-to-speech (TTS) has made significant progress, most existing research remains limited to utterance-level emotional expression and fails to support word-level control. Achieving word-level expressive control poses fundamental challenges, primarily due to the complexity of modeling multi-emotion transitions and the scarcity of annotated datasets that capture intra-sentence emotional and prosodic variation. In this paper, we propose WeSCon, the first self-training framework that enables word-level control of both emotion and speaking rate in a pretrained zero-shot TTS model, without relying on datasets containing intra-sentence emotion or speed transitions.Our method introduces a transition-smoothing strategy and a dynamic speed control mechanism to guide the pretrained TTS model in performing word-level expressive synthesis through a multi-round inference process. To further simplify the inference, we incorporate a dynamic emotional attention bias mechanism and fine-tune the model via self-training, thereby activating its ability for word-level expressive control in an end-to-end manner. Experimental results show that WeSCon effectively overcomes data scarcity, achieving state-of-the-art performance in word-level emotional expression control while preserving the strong zero-shot synthesis capabilities of the original TTS model. | spotlight | 2509.24629 | null | null | [] | [] | [] | [
-0.034847863018512726,
-0.02422965131700039,
-0.01837446726858616,
0.04355314373970032,
0.004427883308380842,
0.043744686990976334,
0.05943167209625244,
0.02564854547381401,
-0.008166192099452019,
-0.027652665972709656,
-0.03682875633239746,
0.03189052641391754,
-0.03390945494174957,
-0.00... |
98 | Structure Matters: Dynamic Policy Gradient | [
"Sara Klein",
"Xiangyuan Zhang",
"Tamer Basar",
"Simon Weissmann",
"Leif Döring"
] | In this work, we study $\gamma$-discounted infinite-horizon tabular Markov decision processes (MDPs) and introduce a framework called dynamic policy gradient (DynPG). The framework directly integrates dynamic programming with (any) policy gradient method, explicitly leveraging the Markovian property of the environment. DynPG dynamically adjusts the problem horizon during training, decomposing the original infinite-horizon MDP into a sequence of contextual bandit problems. By iteratively solving these contextual bandits, DynPG converges to the stationary optimal policy of the infinite-horizon MDP. To demonstrate the power of DynPG, we establish its non-asymptotic global convergence rate under the tabular softmax parametrization, focusing on the dependencies on salient but essential parameters of the MDP. By combining classical arguments from dynamic programming with more recent convergence arguments of policy gradient schemes, we prove that softmax DynPG scales polynomially in the effective horizon $(1-\gamma)^{-1}$. Our findings contrast recent exponential lower bound examples for vanilla policy gradient. | poster | 2411.04913 | null | null | [] | [] | [] | [
-0.04975588992238045,
-0.015926970168948174,
0.004036898259073496,
0.05035332217812538,
0.04029364511370659,
0.02627614326775074,
0.01920373924076557,
0.007028238847851753,
-0.043934907764196396,
-0.04496666416525841,
-0.021216051653027534,
0.016338909044861794,
-0.07223210483789444,
-0.01... |
99 | Spectral Compressive Imaging via Chromaticity-Intensity Decomposition | [
"Xiaodong Wang",
"Zijun He",
"Ping Wang",
"Lishun Wang",
"Yanan Hu",
"Xin Yuan"
] | In coded aperture snapshot spectral imaging (CASSI), the captured measurement entangles spatial and spectral information, posing a severely ill-posed inverse problem for hyperspectral images (HSIs) reconstruction. Moreover, the captured radiance inherently depends on scene illumination, making it difficult to recover the intrinsic spectral reflectance that remains invariant to lighting conditions. To address these challenges, we propose a \textbf{chromaticity-intensity decomposition framework}, which disentangles an HSI into a spatially smooth intensity map and a spectrally variant chromaticity cube. The chromaticity encodes lighting-invariant reflectance, enriched with high-frequency spatial details and local spectral sparsity. Building on this decomposition, we develop \textbf{CIDNet}—a Chromaticity-Intensity Decomposition unfolding network within a dual-camera CASSI system. CIDNet integrates a hybrid spatial-spectral Transformer tailored to reconstruct fine-grained and sparse spectral chromaticity and a degradation-aware, spatially-adaptive noise estimation module that captures anisotropic noise across iterative stages. Extensive experiments on both synthetic and real-world CASSI datasets demonstrate that our method achieves superior performance in both spectral and chromaticity fidelity. Code and models will be publicly available. | poster | 2509.16690 | null | null | [] | [] | [] | [
-0.009597166441380978,
0.011728104203939438,
-0.04019852355122566,
0.041608043015003204,
0.045289549976587296,
0.03266443312168121,
0.016881728544831276,
0.006966606248170137,
-0.048430923372507095,
-0.0820155143737793,
0.002285463036969304,
-0.011517174541950226,
-0.045795369893312454,
-0... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.