AAAI 2026 Papers — Page 17
AAAI Conference on Artificial Intelligence · 4149 papers
GOAL: Geometrically Optimal Alignment for Continual Generalized Category Discovery
Jizhou Han (Xi'an Jiaotong University), Yihong Gong (Harbin Institute of Technology)
ClassificationRepresentation LearningContrastive LearningImage
🎯 What it does: Proposes a unified framework called GOAL based on a fixed equiangular tight frame (ETF) classifier to address the issues of forgetting and class confusion in continual generalized class discovery (C-GCD).
GOMPSNR: Reflourish the Signal-to-Noise Ratio Metric for Audio Generation Tasks
Lingling Dai (Institute of Acoustics, Chinese Academy of Sciences), Chengshi Zheng (Institute of Acoustics, Chinese Academy of Sciences)
GenerationMultimodalityAudio
🎯 What it does: Propose an improved audio quality evaluation metric called GOMPSNR, and design new phase-guided and joint amplitude-phase optimization loss functions based on this metric to enhance the generation quality of neural vocoders and audio codecs.
Good Gradients Poison Your Model: Evading Defenses in Federated Learning via Boundary-adaptive Perturbation
Xiaojie Zhao (Beijing University of Posts and Telecommunications), Chongru Fan (Beijing University of Posts and Telecommunications)
Federated LearningImage
🎯 What it does: Propose BAPerturb, a model poisoning attack based on gradient boundaries, which can bypass various defense methods in federated learning.
Good-for-MDP State Reduction for Stochastic LTL Planning
Christoph Weinhuber (University of Oxford), Qiyi Tang (University of Liverpool)
OptimizationReinforcement LearningTextGraphBenchmark
🎯 What it does: Proposes a novel GFM automaton state space reduction technique for Markov Decision Process (MDP) planning targeting Linear Temporal Logic (LTL) objectives, and provides a direct GFM construction method for GF φ structures;
Gotta Hear Them All: Towards Sound Source Aware Audio Generation
Wei Guo (University of Sydney), Weidong Cai (University of Sydney)
GenerationTransformerDiffusion modelAuto EncoderContrastive LearningImageVideoTextMultimodalityAudio
🎯 What it does: Developed an audio generation framework named SS2A, capable of perceiving and fusing multiple sound sources based on three modalities—visual, text, and audio—to achieve more immersive and controllable audio generation.
GP-MoLFormer-Sim: Test Time Molecular Optimization Through Contextual Similarity Guidance
Jiří Navrátil (IBM Research), Brian Belgodere (IBM Research)
OptimizationDrug DiscoveryTransformerLarge Language ModelTextBiomedical DataBenchmark
🎯 What it does: Propose a training-agnostic test-time molecular generation method, GP-MOLFORMER-SIM, which leverages context similarity during the generation process to guide a chemical language model in generating new molecules similar to the target molecule, and integrates it into a genetic algorithm to improve optimization efficiency.
GPGS: Consistent 3D Object Removal via Geometry-Aware 3D Inpainting and Projected Image Refinement in 3D Gaussian Splatting
Yongjoon Lee (Hanyang University), Donghyeon Cho (Hanyang University)
RestorationConvolutional Neural NetworkAuto EncoderGaussian SplattingImagePoint Cloud
🎯 What it does: By first separating the target object within the 3D Gaussian Splatting framework, using Point-MAE to complete the geometry of unobserved regions, then performing brightness and texture correction on reference images projected to different viewpoints, and finally fine-tuning 3DGS with the corrected images, achieving multi-view consistent 3D object removal.
Gracefully Air-Written: Enhancing the Legibility and Style Consistency of In-Air Handwriting
Yu Liu (Dalin Minzu University), Bo Lu (Dalin Minzu University)
GenerationTransformerDiffusion modelContrastive LearningImage
🎯 What it does: Proposed a binary encoding scheme based on diffusion models that can reconstruct and enhance the readability and author style consistency of 3D space handwritten characters from a small number of examples.
Gradient as Conditions: Rethinking HOG for All-in-one Image Restoration
Jiawei Wu (Sun Yat-sen University), Zhi Jin (Sun Yat-sen University)
RestorationTransformerImage
🎯 What it does: This paper proposes a Transformer model called HOGformer, which integrates a learnable HOG gradient prior for unified image restoration.
Gradient-Protected Value Decomposition for Cooperative Multi-Agent Reinforcement Learning
Jie Hou (Xi'an Jiaotong University), Chenyang Ge (Xi'an Jiaotong University)
OptimizationReinforcement Learning
🎯 What it does: Propose a new multi-agent reinforcement learning framework GPVD, specifically designed to address gradient interference issues in value decomposition methods, and improve the efficiency of collaborative strategy learning through a gradient protection mechanism.
GraFT: Infusing Pre-trained Transformers with Relational Structure for Time Series Forecasting
Yuqi Yuan (University of Science and Technology Beijing), Wenbing Zhao (Cleveland State University)
Graph Neural NetworkTransformerLarge Language ModelSupervised Fine-TuningTime SeriesBenchmark
🎯 What it does: Improving long-term prediction for multivariate time series by constructing a Heterogeneous Patch Relationship Graph (HPRG) and using R-GCN to inject structural priors into pre-trained Transformer models.
GRAM-R²: Self-Training Generative Foundation Reward Models for Reward Reasoning
Chenglong Wang (Northeastern University), Tong Xiao (Northeastern University)
GenerationData-Centric LearningReinforcement Learning from Human FeedbackTransformerLarge Language ModelSupervised Fine-TuningReinforcement LearningTextChain-of-Thought
🎯 What it does: Designed and trained a self-supervised pre-trained generative reward model, GRAM-R², which enhances reward reasoning capabilities through self-training iterations using unlabeled data, and provides a dedicated preference proof model to generate reasoning processes for unlabeled labels.
GranAlign: Granularity-Aware Alignment Framework for Zero-shot Video Moment Retrieval
Mingyu Jeon (Chung-Ang University), Junyeong Kim (Electronics and Telecommunications Research Institute)
RetrievalTransformerLarge Language ModelVision Language ModelContrastive LearningVideoText
🎯 What it does: Propose a zero-shot video moment retrieval framework GranAlign, achieving semantic matching through multi-grained query rewriting and query-aware caption generation;
Graph Attention-Guided Search for Dense Multi-Agent Pathfinding
Rishabh Jain (University of Cambridge), Amanda Prorok (University of Cambridge)
OptimizationGraph Neural NetworkSupervised Fine-TuningGraph
🎯 What it does: Propose a hybrid multi-agent path planning framework (LaGAT) that uses graph attention neural networks as heuristic guidance for search.
Graph Choosability via SAT: Beyond the Nullstellensatz
Markus Kirchweger (TU Wien), Stefan Szeider (TU Wien)
OptimizationGraph
🎯 What it does: This paper employs a hybrid method combining SAT and QBF to construct a custom propagator for solving the optional graph coloring problem.
Graph Contrastive Learning with Balanced Hard Negatives and Fine-grained Semantic-aware Positives
Hongshan Pu (South China University of Technology), Hongmin Cai (South China University of Technology)
ClassificationRepresentation LearningAdversarial AttackGraph Neural NetworkContrastive LearningGraph
🎯 What it does: Propose the BalanceGCL framework, which improves the representation quality of unsupervised graph contrastive learning by generating class-balanced hard negative samples and fine-grained semantic positive samples.
Graph Domain Adaptation via Homophily-Agnostic Reconstructing Structure
Ruiyi Fang, Boyu Wang (Western University)
Domain AdaptationGraph Neural NetworkGraph
🎯 What it does: This paper proposes an unsupervised graph domain adaptation framework called RSGDA, which reconstructs two types of graph structures (intra-class affinity and inter-class affinity), and achieves label-free transfer between source and target graphs using adaptive filters and an alignment network.
Graph Flow Matching: Enhancing Image Generation with Neighbor-Aware Flow Fields
Md Shahriar Rahim Siddiqui, Eldad Haber (University of British Columbia)
GenerationGraph Neural NetworkFlow-based ModelImageOrdinary Differential Equation
🎯 What it does: Propose Graph Flow Matching (GFM), integrating graph neural networks into the flow matching framework to achieve neighbor-aware velocity correction, thereby improving image generation quality.
Graph Masked Autoencoder for Multi-view Remote Sensing Data Clustering
Renxiang Guan (National University of Defense Technology), Xinwang Liu (Changsha University)
Representation LearningGraph Neural NetworkAuto EncoderContrastive LearningImageGraph
🎯 What it does: Proposed a graph masked autoencoder framework named CG-MAE for clustering multi-view remote sensing data.
Graph Meets Deep Unfolding: An Interpretable Mutual-benefit Multi-view Learning Network
Renjie Lin (University of Electronic Science and Technology of China), Le Zhang (University of Electronic Science and Technology of China)
OptimizationExplainability and InterpretabilityRepresentation LearningGraph Neural NetworkGraph
🎯 What it does: Designed an interpretable deep unrolling network, IMML-Net, for multi-view graph learning, integrating sparse, low-rank constraints, and noise handling to achieve joint learning of multi-view graphs.
Graph Neural Field with Spatial-Correlation Augmentation for HRTF Personalization
De Hu (Inner Mongolia University), Cuicui Jiang (Inner Mongolia University)
GenerationGraph Neural NetworkAudio
🎯 What it does: Propose the GraphNF-SCA framework, which first generates position-agnostic personalized HRTF using graph neural networks, and then refines it through a spatial correlation graph network to achieve high-quality HRTF prediction for unseen subjects.
Graph of Verification: Structured Verification of LLM Reasoning with Directed Acyclic Graphs
Jiwei Fang (Shandong University), Zhiwei Xu (Shandong University)
Explainability and InterpretabilityLarge Language ModelTextBenchmark
🎯 What it does: Propose the Graph of Verification (GoV) framework, which uses directed acyclic graphs (DAGs) and configurable node blocks to perform hierarchical step-by-step verification of the LLM reasoning process.
Graph Out-of-Distribution Detection via Test-Time Calibration with Dual Dynamic Dictionaries
Yue Hou (Beihang University), Ke Xu (Guangxi Normal University)
Anomaly DetectionRepresentation LearningGraph Neural NetworkGraph
🎯 What it does: Proposes a boundary-aware calibration method called BaCa based on dual dynamic dictionaries, achieving test-time OOD detection for graph data without fine-tuning pre-trained graph neural networks or using auxiliary OOD samples.
Graph Smoothing for Enhanced Local Geometry Learning in Point Cloud Analysis
Shangbo Yuan, Na Zhao (University Of Electronic Science And Technology Of China)
ClassificationSegmentationRepresentation LearningGraph Neural NetworkPoint Cloud
🎯 What it does: This paper proposes a point cloud analysis framework named GSPoint, combining a graph smoothing module and a local geometric learning module to address issues caused by traditional spherical query methods, such as sparse boundary connections and cross-node noise;
Graph VQ-Transformer (GVT): Fast and Accurate Molecular Generation via High-Fidelity Discrete Latents
Haozhuo Zheng (Harbin Institute of Technology), Yang Liu (Harbin Institute of Technology)
Drug DiscoveryGraph Neural NetworkTransformerAuto EncoderGraphBenchmark
🎯 What it does: Proposed the Graph VQ-Transformer (GVT) two-stage generation framework: the first stage uses Graph VQ-VAE to compress molecular graphs into high-fidelity discrete latent sequences; the second stage trains a self-attention Transformer to generate new molecules based on these latent sequences.
Graph-augmented and Over-smoothing-resistant Contrastive Clustering for Short Text
Zijian Zheng (Sun Yat-sen University), Yonghe Lu (Sun Yat-sen University)
Representation LearningGraph Neural NetworkTransformerContrastive LearningTextBenchmark
🎯 What it does: Propose the GOCC framework, which utilizes sentence-level and cluster-level graph structures to enhance representation learning and contrastive learning for short text unsupervised clustering.
Graph-Conditional Flow Matching for Relational Data Generation
Davide Scassola (University of Trieste), Luca Bortolussi (Aindo)
GenerationData SynthesisGraph Neural NetworkFlow-based ModelTabularOrdinary Differential Equation
🎯 What it does: This paper proposes a generative model based on graph conditional flow matching, which encodes the foreign key graph of relational databases using graph neural networks to generate all table data content in an end-to-end manner in one go.
Graph-Driven Domain Co-Adaptation for Cross-Domain Image Quality Assessment
Shun Zhu (Nanjing Normal University), Xiaobo Shen (Nanjing University of Science and Technology)
Domain AdaptationConvolutional Neural NetworkGraph Neural NetworkTransformerImage
🎯 What it does: This paper proposes a graph-based domain collaborative adaptation framework called GDCIQA for cross-domain image quality assessment.
Graph-of-Mark: Promote Spatial Reasoning in Multimodal Language Models with Graph-Based Visual Prompting
Giacomo Frisoni (University of Bologna), Gianluca Moro (University of Bologna)
Object DetectionSegmentationDepth EstimationRepresentation LearningPrompt EngineeringVision Language ModelImageTextGraph
🎯 What it does: This paper proposes Graph-of-Mark (GoM), a pixel-level visual prompting technique that overlays scene graphs (nodes as detected objects, edges as their spatial relationships) on images to enhance the zero-shot performance of multimodal language models in spatial reasoning tasks.
Graph-Semantic Guided Learning for Virtual Immunohistochemistry Staining on Consecutive Histology Sections
Fanhao Qiu (Hainan University), Zhengxia Wang (Hainan University)
Image TranslationGenerationGraph Neural NetworkGenerative Adversarial NetworkContrastive LearningImageGraphBiomedical Data
🎯 What it does: Propose a graph semantic guided virtual immunohistochemistry staining framework, GSGStain, which converts images into cell graphs to address spatial mismatch and semantic noise between adjacent slices, generating high-quality, pathology-consistent IHC images.
Graph2Video: Leveraging Video Models to Model Dynamic Graph Evolution
Hua Liu (Southern University of Science and Technology), Yu Zhang (City University of Hong Kong)
Graph Neural NetworkTransformerVideoGraph
🎯 What it does: This paper proposes the Graph2Video framework, which serializes the spatial-temporal neighborhood of target edges in dynamic graphs into 'graph videos,' and extracts spatiotemporal embeddings using a frozen video base model as a lightweight, plug-and-play edge-level memory unit to enhance dynamic link prediction performance.
GraphCoT-VLA: A 3D Spatial-Aware Reasoning Vision-Language-Action Model for Robotic Manipulation with Ambiguous Instructions
Helong Huang (Huawei), Hong Zhang (Huawei)
Robotic IntelligenceGraph Neural NetworkMixture of ExpertsVision Language ModelVision-Language-Action ModelFlow-based ModelImageTextGraphChain-of-Thought
🎯 What it does: Proposed an end-to-end GraphCoT-VLA model capable of 3D spatial perception and action planning when robots encounter ambiguous instructions and open environments;
GraphGrasp: Lightweight and Efficient Graph-Guided 6-DoF Robotic Grasp Pose Estimation Network
Sheng Yu, Yuanqing Xia (Beijing Institute Of Technology)
Pose EstimationRobotic IntelligenceGraph Neural NetworkTransformerPoint Cloud
🎯 What it does: Proposed a lightweight graph-guided 6-DoF grasping pose estimation network called GraphGrasp, which can directly construct scene graphs, object graphs, and grasp graphs from point clouds to achieve grasping point prediction and pose reasoning.
GraphIC: A Graph-Based In-Context Example Retrieval Model for Multi-Step Reasoning
Jiale Fu (Southeast University), Xu Yang (Southeast University)
RetrievalGraph Neural NetworkLarge Language ModelGraph
🎯 What it does: Propose a graph-based thinking graph model called GraphIC to enhance context example retrieval in multi-step reasoning tasks.
GraphIF: Enhancing Multi-Turn Instruction Following for Large Language Models with Relation Graph Prompt
Zhenhe Li (University of Science and Technology of China), Qi Song (University of Science and Technology of China)
TransformerLarge Language ModelSupervised Fine-TuningAgentic AIPrompt EngineeringTextBenchmark
🎯 What it does: Proposed the GraphIF framework, which utilizes relational graphs to model multi-turn dialogues and generate graph prompts, enhancing the multi-turn instruction following capability of large language models.
GraphOracle: Efficient Fully-Inductive Knowledge Graph Reasoning via Relation-Dependency Graphs
Enjun Du (Hong Kong University of Science and Technology (Guangzhou)), Yongqi Zhang (Hong Kong University of Science and Technology (Guangzhou))
Computational EfficiencyRepresentation LearningAdversarial AttackGraph Neural NetworkSupervised Fine-TuningGraph
🎯 What it does: Propose a novel knowledge graph reasoning framework named GRAPHORACLE based on a relation dependency graph (RDG), which achieves efficient and accurate chain-of-thought reasoning in fully inductive (both entities and relations unseen) and cross-domain reasoning scenarios.
GraphRAG-Induced Dual Knowledge Structure Graphs for Personalized Learning Path Recommendation
Xinghe Cheng (Jinan University), Weiqi Luo (Griffith University)
Recommendation SystemGraph Neural NetworkLarge Language ModelReinforcement LearningAgentic AITextGraphRetrieval-Augmented Generation
🎯 What it does: Propose the KnowLP framework, which combines prerequisite relationships and similar relationships to construct a dual knowledge structure graph, using EDU-GraphRAG to automatically generate knowledge graphs and DLRL (including three agents: prerequisite, similar, and difficulty) to generate personalized learning paths.
GraphTextack: A Realistic Black-Box Node Injection Attack on LLM-Enhanced GNNs
Jiaji Ma (University of Michigan), Danai Koutra (University of Michigan)
Adversarial AttackGraph Neural NetworkLarge Language ModelMultimodality
🎯 What it does: Proposes a black-box multi-modal node injection attack method called GRAPHTEXTACK for LLM-enhanced GNNs
GrayKD: Distilling Better Knowledge from Black-box LLM via Multi-rationale Injection
Hyeongsoo Lim (Chung-Ang University), Ji Won Yoon (42dot Inc)
Knowledge DistillationTransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringText
🎯 What it does: Developed a GrayKD framework that uses a single-stage approach to extract text-level knowledge from black-box LLMs and perform knowledge distillation.
GRDC: A Unified Graph-Driven Framework for Role Discovery and Communication in Multi-Agent Reinforcement Learning
Zihong Gao (Xi'an Jiaotong University), Liangjun Ke (Xi'an Jiaotong University)
Graph Neural NetworkReinforcement LearningGraph
🎯 What it does: Propose the GRDC framework, which achieves multi-agent coordination under partial observability by constructing a local interaction graph for role discovery and intra-role communication.
Greedily Maximizing Ex-Ante Fairness
Ruben Becker (Ca Foscari University Of Venice), Cosimo Vinci (University Of Salento)
Optimization
🎯 What it does: Propose a general framework based on ex-ante fairness, and provide greedy algorithms for three types of constraints (cardinality, arbitrary sets, and allocation);
Griffin: Aerial-Ground Cooperative Detection and Tracking Dataset and Benchmark
Jiahao Wang (Tsinghua University), Jianqiang Wang (Tsinghua University)
Object DetectionObject TrackingAutonomous DrivingTransformerVideoBenchmark
🎯 What it does: Constructed and released the Griffin dataset and the corresponding benchmark for UAV-ground collaborative 3D perception tasks.
GRIM: Task-Oriented Grasping with Conditioning on Generative Examples
Shailesh (IIT Dhanbad), Gora Chand Nandi (University of Bremen)
Robotic IntelligenceVision-Language-Action ModelImageVideoPoint CloudRetrieval-Augmented Generation
🎯 What it does: Propose a training-agnostic task-oriented grasping framework GRIM, which retrieves, aligns, and transfers functional grasping poses from multi-source memories through a retrieval-alignment-transfer process.
GRIP: Latent Field-Guided Graph Policy for Budget-Constrained Multi-Agent Routing
Yujiao Hu (Chang'an University), Yan Pan (National University of Defense Technology)
OptimizationGraph Neural NetworkReinforcement LearningGraph
🎯 What it does: Proposed and implemented the GRIP framework to address subset selection and path planning problems for multi-agent systems under budget constraints.
Ground What You See: Hallucination-Resistant MLLMs via Caption Feedback, Diversity-Aware Sampling, and Conflict Regularization
Miao Pan (Zhejiang University), Xuhong Zhang (National Certification Technology Co Ltd)
Explainability and InterpretabilityReinforcement Learning from Human FeedbackTransformerLarge Language ModelReinforcement LearningVision Language ModelContrastive LearningVideoTextMultimodality
🎯 What it does: Propose an RL optimization framework based on visual calibration, reward variance sampling, and conflict regularization, significantly reducing hallucinations in multimodal large language models.
Grounding Actions in Camera Space: Observation-Centric Vision-Language-Action Policy
Tianyi Zhang (Zhejiang University), Zhi Hou (Shanghai Artificial Intelligence Laboratory)
Robotic IntelligenceTransformerVision Language ModelVision-Language-Action ModelDiffusion modelImageMultimodality
🎯 What it does: Proposes an OCVLA framework that aligns robot action prediction to camera space, enabling direct prediction of actions in the observation space and addressing the mismatch between perception and action spaces.
Group Causal Policy Optimization for Post-Training Large Language Models
Ziyin Gu (Institute of Software Chinese Academy of Sciences), Wenwen Qiang (Institute of Software Chinese Academy of Sciences)
OptimizationTransformerLarge Language ModelReinforcement LearningText
🎯 What it does: Proposes the Group Causal Policy Optimization (GCPO) algorithm, which enhances the reasoning and generation performance of large language models through causal-structured group policy optimization during the post-training phase.
Group Fair Matchings Using Convex Cost Functions
Atasi Panda (Indian Institute of Science), Prajakta Nimbhorkar (Chennai Mathematical Institute)
OptimizationFlow-based ModelTabular
🎯 What it does: Proposes the problem of project-platform matching on bipartite graphs, achieving softened fairness constraints by introducing overall and group-specific convex cost functions, aiming to minimize total cost while meeting a given utility threshold.
Group Orthogonal Low-Rank Adaptation for RGB-T Tracking
Zekai Shao (University of Science and Technology Beijing), Hongmin Liu (University of Science and Technology Beijing)
Object TrackingTransformerSupervised Fine-TuningVideo
🎯 What it does: This paper proposes a parameter-efficient fine-tuning framework based on Group Orthogonal Low-Rank Adaptation (GOLA), specifically designed for RGB-T visual object tracking. The framework decomposes the LoRA low-rank matrix via singular value decomposition (SVD), separating key ranks and redundant ranks, and then clusters the redundant ranks into groups. Subsequently, orthogonal constraints are applied across different groups, forcing them to learn complementary features, thereby fully activating the redundant rank space and enhancing model expressiveness.
Group-aware Multiscale Ensemble Learning for Test-Time Multimodal Sentiment Analysis
Kai Tang (Zhejiang University), Haobo Wang (Ant Group)
ClassificationDomain AdaptationContrastive LearningMultimodality
🎯 What it does: Propose the GMEL framework to address the adaptability issue in multi-modal sentiment analysis under test-time domain drift
GROVER: Graph-guided Representation of Omics and Vision with Expert Regulation for Adaptive Spatial Multi-omics Fusion
Yongjun Xiao (Great Bay University), Xubin Zheng (Great Bay University)
Representation LearningGraph Neural NetworkMixture of ExpertsContrastive LearningMultimodalityBiomedical Data
🎯 What it does: Propose the GROVER framework, which performs single-point adaptive fusion of spatial transcriptomics, proteomics, and tissue imaging based on graph convolutional networks, addressing challenges of multi-modal heterogeneity and spatial alignment.
Grow-on-Demand: Sparse and Adaptive Expert Expansion for Continual Instruction Tuning
Ying Zhang (Nankai University), Xiaojie Yuan (Tiangong University)
Computational EfficiencyMeta LearningLarge Language ModelSupervised Fine-TuningMixture of ExpertsText
🎯 What it does: Propose a sparse adaptive expert expansion framework named GoD-MoE, which enables continuous instruction fine-tuning of large language models without data replay, thereby mitigating catastrophic forgetting.
GS-Checker: Tampering Localization for 3D Gaussian Splatting
Haoliang Han (Hong Kong Baptist University), Renjie Wan (Hong Kong Baptist University)
SegmentationAnomaly DetectionContrastive LearningGaussian SplattingPoint Cloud
🎯 What it does: Proposes the GS-Checker method for precisely locating tampered regions in 3D Gaussian Splatting (3DGS) models.
GSAP-ERE: Fine-Grained Scholarly Entity and Relation Extraction Focused on Machine Learning
Wolfgang Otto (GESIS Leibniz Institute for the Social Sciences), Stefan Dietze (GESIS Leibniz Institute for the Social Sciences)
TransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringTextBenchmark
🎯 What it does: This paper constructs the GSAP-ERE dataset and evaluates various entity and relation extraction models based on it;
GT-SNT: A Linear-Time Transformer for Large-Scale Graphs via Spiking Node Tokenization
Huizhe Zhang (Sun Yat Sen University), Liang Chen (Sun Yat Sen University)
ClassificationSpiking Neural NetworkTransformerGraph
🎯 What it does: Propose a linear-time graph Transformer called GT-SNT, achieving node classification through Spiking Node Tokenization.
GT2-GS: Geometry-aware Texture Transfer for Gaussian Splatting
Wenjie Liu (East China Normal University), Yang Li (East China Normal University)
Image TranslationGenerationGaussian SplattingImagePoint Cloud
🎯 What it does: Proposed a geometry-aware texture transfer framework called GT-GS aimed at effectively transferring 2D textures to complex 3D scenes.
Guess or Recall? Training CNNs to Classify and Localize Memorization in LLMs
Jérémie Dentan (École Polytechnique), Sonia Vanier (École Polytechnique)
ClassificationExplainability and InterpretabilityConvolutional Neural NetworkLarge Language ModelText
🎯 What it does: An evaluation framework based on convolutional neural networks (CNN) for classifying attention weights is proposed to address the character-by-character memorization phenomenon in large language models, with a redesigned classification system for memory samples based on this framework;
GUI-Eyes: Tool-Augmented Perception for Visual Grounding in GUI Agents
Chen Chen (University of Science and Technology of China), Wu Liu (University of Science and Technology of China)
TransformerLarge Language ModelSupervised Fine-TuningReinforcement LearningVision-Language-Action ModelMultimodality
🎯 What it does: Proposed the GUI-Eyes framework, enabling GUI agents to proactively decide when and how to invoke visual tools such as cropping and zooming during inference, thereby achieving a closed-loop of dynamic perception and decision-making.
GUI-G²: Gaussian Reward Modeling for GUI Grounding
Fei Tang (Zhejiang University), Yueting Zhuang (Zhejiang University)
OptimizationTransformerReinforcement LearningVision Language ModelGaussian SplattingMultimodality
🎯 What it does: Designed the GUI-G2 reward framework, modeling GUI elements with a continuous 2D Gaussian distribution to achieve denser learning signals.
GUIC: Certified Graph Unlearning with Individual Fairness Guarantees
Zichong Wang (Florida International University), Wenbin Zhang (Florida International University)
ClassificationGraph Neural NetworkGraph
🎯 What it does: Proposed the GUIC framework to achieve individual fairness in graph models while satisfying the 'forgetting' requirement.
GUIDE: Gaussian Unified Instance Detection for Enhanced Obstacle Perception in Autonomous Driving
Chunyong Hu (Alibaba Group), Sheng Yang (Alibaba Group)
Object DetectionObject TrackingAutonomous DrivingTransformerGaussian SplattingPoint Cloud
🎯 What it does: Developed the GUIDE framework, leveraging 3D Gaussian sparse representation to achieve instance-level occupancy prediction, and performing object detection and tracking within the same framework.
Guided Distillation and Risk Adaptive Evolution for Multi-Robot Navigation
Xuyang Li (Xi'an Jiaotong University), Jianru Xue (Xi'an Jiaotong University)
OptimizationKnowledge DistillationRobotic IntelligenceTransformerLarge Language ModelReinforcement LearningPrompt EngineeringTextChain-of-Thought
🎯 What it does: This paper proposes a two-phase offline reinforcement learning framework called GUIDER, which first utilizes a large language model (LLM) to distill knowledge into a lightweight policy guidance model (PGM), providing reliable guidance when the agent is uncertain; subsequently, the LLM is transformed into a semantic evolution engine to automatically search and generate environment-specific risk-adaptive strategies, achieving safe and efficient multi-robot navigation.
Guided Perturbation Sensitivity (GPS): Detecting Adversarial Text via Embedding Stability and Word Importance
Bryan E. Tuck (University of Houston), Rakesh M. Verma (University of Houston)
Adversarial AttackRecurrent Neural NetworkTransformerText
🎯 What it does: Propose the Guided Perturbation Sensitivity (GPS) framework, which detects adversarial text by masking important words and measuring embedding changes.
GuideGen: A Text-Guided Framework for Paired Full-torso Anatomy and CT Volume Generation
Linrui Dai (Shanghai Jiao Tong University), Xiaofan Zhang (Shanghai Jiao Tong University)
SegmentationGenerationData SynthesisConvolutional Neural NetworkDiffusion modelGenerative Adversarial NetworkTextBiomedical DataComputed Tomography
🎯 What it does: Developed GuideGen, a controllable generation framework capable of generating full-body CT volumes and corresponding anatomical masks using only structured medical text prompts.
Guideline-Consistent Segmentation via Multi-Agent Refinement
Vanshika Vats (University of California, Santa Cruz), James Davis (University of California, Santa Cruz)
SegmentationReinforcement LearningAgentic AIVision Language ModelImageTextRetrieval-Augmented Generation
🎯 What it does: Propose a training-agnostic multi-agent framework that strictly adheres to long-form, fine-grained annotation guidelines for semantic segmentation results through Worker–Supervisor iterative loops and RL adaptive stopping strategies.
GUIDER: Uncertainty Guided Dynamic Re-ranking for Large Language Models Based Recommender Systems
Cai Xu (Xidian University), Meng Yan (Xidian University)
Recommendation SystemTransformerLarge Language ModelTextSequential
🎯 What it does: Design and implement the GUIDER framework to quantify and decompose uncertainty in LLM-generated recommendations, and enhance recommendation reliability through four-quadrant dynamic re-ranking.
Guiding Point Cloud Denoising with Learned Structural Priors
Chuchen Guo, Ying He (Nanyang Technological University)
RestorationConvolutional Neural NetworkAuto EncoderPoint Cloud
🎯 What it does: Propose a point cloud denoising framework based on structural prior guidance, first extracting local geometric structural priors through vector quantization, and then achieving fine-grained feature reconstruction using FiLM-modulated attention.
GuidNoise: Single-Pair Guided Diffusion for Generalized Noise Synthesis
Changjin Kim (SNUAILAB), YoungJoon Yoo (SNUAILAB)
RestorationData SynthesisConvolutional Neural NetworkDiffusion modelScore-based ModelImage
🎯 What it does: Generate realistic noise using a single-pair guided diffusion model, achieving noise synthesis without camera metadata and requiring only a pair of noisy/clean images
GUSLO: General and Unified Structured Light Optimization
Tinglei Wan (Harbin Institute of Technology), Tonghua Su (Harbin Institute of Technology)
Depth EstimationOptimizationImagePoint CloudBenchmark
🎯 What it does: Proposed a generic unified structured light optimization framework, GUSLO, which can accomplish geometric calibration and illumination compensation under a single projection, achieving high-precision 3D reconstruction for various structured light patterns, including binary stripes, speckles, and color-encoded patterns;
H-GAR: A Hierarchical Interaction Framework via Goal-Driven Observation-Action Refinement for Robotic Manipulation
Yijie Zhu (Harbin Institute of Technology), Zitong Yu (Great Bay University)
Robotic IntelligenceTransformerVision-Language-Action ModelDiffusion modelImageTextMultimodality
🎯 What it does: Proposes the H-GAR (Hierarchical Interaction Framework via Goal-Driven Observation-Action Refinement) framework, which first predicts the task goal image and generates coarse-grained actions, then uses a goal-conditioned observation synthesizer (GOS) to generate intermediate observations, followed by an interaction-aware action refiner (IAAR) that refines actions by combining historical action memory, thereby achieving end-to-end robotic manipulation planning.
H-RDT: Human Manipulation Enhanced Bimanual Robotic Manipulation
Hongzhe Bi (Tsinghua University), Jun Zhu (Horizon Robotics)
Representation LearningRobotic IntelligenceTransformerSupervised Fine-TuningDiffusion modelFlow-based ModelVideoText
🎯 What it does: Pretrain a robot control model using 3D hand pose data from large-scale self-captured human-operated videos, then perform cross-body fine-tuning through modular action encoding/decoding to achieve dual-arm robot manipulation;
HABIT: Chrono-Synergia Robust Progressive Learning Framework for Composed Image Retrieval
Zixu Li (Shandong University), Yinwei Wei (Shandong University)
RetrievalTransformerVision Language ModelContrastive LearningMultimodality
🎯 What it does: Propose a robust learning framework named HABIT for addressing noisy triplet correspondence issues in compositional image retrieval, integrating mutual information transfer rate evaluation with dual consistency evolution learning;
Hallucinate Less by Thinking More: Aspect-Based Causal Abstention for Large Language Models
Vy Nguyen (RMIT University), Xiuzhen Zhang (RMIT University)
Explainability and InterpretabilityComputational EfficiencyLarge Language ModelPrompt EngineeringTextBenchmark
🎯 What it does: Proposed and implemented the Aspect-Based Causal Abstention (ABCA) framework, which can decide whether to refuse answering by analyzing the diversity of internal knowledge in LLMs before generation, thereby reducing hallucinations.
Hallucination as a Computational Boundary: A Hierarchy of Inevitability and the Oracle Escape
Xi Wang (Hefei Institutes of Physical Science, Chinese Academy of Sciences), Xianjun Yang (iFLYTEK Research)
Safty and PrivacyExplainability and InterpretabilityComputational EfficiencyTransformerLarge Language ModelTextBenchmarkRetrieval-Augmented Generation
🎯 What it does: Proposes the computer science nature of hallucinations, clarifying the origins of LLM hallucinations by constructing a three-tier hierarchy of inevitability (diagonalization, uncomputability, information theory), and proposes two escape paths: one is achieving absolute escape through an external Oracle (e.g., RAG), and the other is achieving internalized escape through continual learning (CL). Subsequently, it proposes the Computational Class Alignment (CCA) safety principle and verifies the performance of RAG, CLM, and the hybrid RAG-CL in experiments.
HalluClean: A Unified Framework to Combat Hallucinations in LLMs
Yaxin Zhao (Harbin Institute of Technology), Yu Zhang (Harbin Institute of Technology)
Explainability and InterpretabilityTransformerPrompt EngineeringTextBenchmarkFinance RelatedChain-of-Thought
🎯 What it does: Proposes the hallucination removal framework HalluClean, which detects and corrects hallucinations in LLM-generated text using structured reasoning.
HALO: Hardware-Aware Quantization with Low Critical-Path-Delay Weights for LLM Acceleration
Rohan Juneja (National University of Singapore), Li-Shiuan Peh (National University of Singapore)
Computational EfficiencyTransformerText
🎯 What it does: This paper proposes HALO, a hardware-aware post-training quantization framework that efficiently accelerates LLM inference by combining weight quantization with dynamic voltage and frequency scaling (DVFS).
HALoRA: Low-Rank Adaptation with Hierarchical Budget Allocation for Efficient Vision-Language Alignment
Letian Zhang (Tsinghua University), Jinpeng Wang (Tsinghua University)
ClassificationRetrievalComputational EfficiencyTransformerVision Language ModelImageVideoTextMultimodality
🎯 What it does: Propose HALoRA, a hierarchical low-rank adaptation method for dual-encoder vision-language alignment;
HAMLET4Fairness: Enhancing Fairness in AI Pipelines Through Human-Centered AutoML and Argumentation
Joseph Giovanelli (Alma Mater Studiorum - University of Bologna), Roberta Calegari (Alma Mater Studiorum - University of Bologna)
OptimizationExplainability and InterpretabilityHyperparameter SearchTabularBenchmark
🎯 What it does: Developed HAMLET4Fairness, a framework that integrates structured argumentation with human-centric AutoML for fairness-driven AI pipeline optimization;
HaNa: Hardness and Noise-Aware Robust Cross-modal Retrieval
Fangming Zhong (Dalian University of Technology), Suhua Zhang (Dalian University of Technology)
RetrievalTransformerVision Language ModelContrastive LearningMultimodality
🎯 What it does: Proposes the HaNa method, which combines an adaptive reweighting mechanism for clean samples and asymmetric regularization for noisy samples to enhance cross-modal retrieval performance under noisy conditions.
HandMCM: Multi-modal Point Cloud-based Correspondence State Space Model for 3D Hand Pose Estimation
Wencan Cheng (National University of Singapore), Gim Hee Lee (National University of Singapore)
Pose EstimationMultimodalityPoint Cloud
🎯 What it does: This paper proposes a multi-modal Mamba architecture called HandMCM for estimating 3D hand keypoint positions from RGB-D inputs.
HanjaBridge: Resolving Semantic Ambiguity in Korean LLMs via Hanja-Augmented Pre-Training
Seungho Choi (Wisenut), Bongsu Kim (Wisenut)
Computational EfficiencyKnowledge DistillationRepresentation LearningTransformerLarge Language ModelContrastive LearningText
🎯 What it does: Propose HanjaBridge, a semantic enhancement technique that inserts Hanja (Chinese characters) candidate words into Korean LLM pre-training processes, leveraging multiple candidates and context to resolve semantic ambiguity in Korean homonyms; based on this, continuous pre-training (CPT) and token-level knowledge distillation are adopted to prevent catastrophic forgetting, and no additional Hanja tokens are required during inference, maintaining original inference efficiency.
HAP: Harmonized Amplitude Perturbation for Cross-Domain Few-Shot Learning
Wenqian Li (Southeast University), Hui Xue (Southeast University)
ClassificationDomain AdaptationImage
🎯 What it does: Proposes the Harmonized Amplitude Perturbation (HAP) frequency-domain augmentation strategy, perturbing the amplitude spectrum to enhance cross-domain few-shot learning performance.
HAPO: Training Language Models to Reason Concisely via History-Aware Policy Optimization
Chengyu Huang (Cornell University), Claire Cardie (Cornell University)
OptimizationComputational EfficiencyReinforcement Learning from Human FeedbackTransformerLarge Language ModelReinforcement LearningTextBenchmark
🎯 What it does: Train LLMs for more concise reasoning by proposing a length reward mechanism based on historical information called HAPO.
Hard vs. Noise: Resolving Hard-Noisy Sample Confusion in Recommender Systems via Large Language Models
Tianrui Song (Hong Kong University of Science and Technology (Guangzhou)), Hao Liu (Hong Kong University of Science and Technology)
Recommendation SystemLarge Language ModelContrastive LearningTextGraph
🎯 What it does: This work proposes the LLMHNI framework, which leverages two auxiliary signals provided by large language models (LLMs)—semantic relevance and logical relevance—to address the confusion between hard samples and noisy samples in recommendation systems. It achieves more robust implicit recommendations through semantic-guided hard negative sampling and logic relevance-driven interaction denoising.
HardF-SNN: Hardware-Friendly Quantization for Spiking Neural Networks with Efficient Integer-Arithmetic-Only Inference
Hanwen Liu (University of Electronic Science and Technology of China), Malu Zhang (University of Electronic Science and Technology of China)
ClassificationSpiking Neural NetworkImage
🎯 What it does: Propose a lightweight, hardware-friendly quantized SNN framework called HardF-SNN that supports integer-only inference.
Harmonic Dataset Distillation for Time Series Forecasting
Seungha Hong (Pohang University of Science and Technology), Hwanjo Yu (University of Illinois Urban-Champaign)
Data SynthesisKnowledge DistillationTime Series
🎯 What it does: Proposes a method for time series distillation in the frequency domain (HDT), which decomposes sequences into sinusoidal bases via FFT and matches principal harmonics to generate a compact synthetic dataset;
HarmoQ: Harmonized Post-Training Quantization for High-Fidelity Image Super-Resolution
Hongjun Wang (University of Tokyo), Yinqiang Zheng (University of Tokyo)
RestorationSuper ResolutionTransformerImage
🎯 What it does: Propose HarmoQ, a unified post-training quantization framework specifically designed for Transformer-based super-resolution models, combining structural residual calibration, quantization scale balance, and adaptive boundary refinement to achieve coordinated optimization of weights and activations.
Harnessing Textual Semantic Priors for Knowledge Transfer and Refinement in CLIP-Driven Continual Learning
Lingfeng He, Nannan Wang (Huawei Technologies Co Ltd)
Knowledge DistillationRepresentation LearningTransformerSupervised Fine-TuningVision Language ModelMultimodality
🎯 What it does: This paper proposes a unified framework called SECA, which leverages text semantic priors to guide knowledge transfer and visual prototype refinement in CLIP-driven continual learning.
Harnessing the Unseen: The Hidden Influence of Intrinsic Knowledge in Long-Context Language Models
Yu Fu (University of California Riverside), Yue Dong (University of California Riverside)
RetrievalTransformerLarge Language ModelTextRetrieval-Augmented Generation
🎯 What it does: Investigate the impact of parameter knowledge (intrinsic knowledge) in long context language models during generation, and propose a hybrid Needle-in-a-Haystack evaluation method to simultaneously assess parameter recall and external retrieval capabilities.
Harnessing Vision-Language Models for Time Series Anomaly Detection
Zelin He (Pennsylvania State University), Matthew Reimherr (Pennsylvania State University)
Anomaly DetectionTransformerVision Language ModelTime Series
🎯 What it does: This paper proposes a two-stage zero-shot time series anomaly detection framework called VLM4TS, which converts one-dimensional sequences into two-dimensional images. It first uses ViT4TS to locate candidate anomalies and then employs VLM4TS to verify precise boundaries.
Hashed Watermark as a Filter: A Unified Defense Against Forging and Overwriting Attacks in Neural Network Watermarking
Yuan Yao (Beijing Teleinfo Technology Company Ltd.), Jian Jin (China Academy of Information and Communications Technology)
Safty and PrivacyConvolutional Neural NetworkTransformerImageText
🎯 What it does: Designed and verified a white-box neural network watermarking method called NeuralMark based on hash watermark filters, which can resist forgery, overwriting, fine-tuning, and pruning attacks.
HATIR: Heat-Aware Diffusion for Turbulent Infrared Video Super-Resolution
Yang Zou (Northwestern Polytechnical University), Jinyuan Liu (Dalian University of Technology)
Super ResolutionDiffusion modelOptical FlowVideo
🎯 What it does: Proposed the HATIR framework to achieve thermal-aware diffusion-based thermal infrared video super-resolution.
HC2-GNN: Hierarchical Graph Representation Learning for Efficient Text Classification
jiejie fan (University of Science and Technology Beijing), Xi Sun (University of Science and Technology Beijing)
ClassificationRepresentation LearningGraph Neural NetworkText
🎯 What it does: This paper proposes HC2-GNN, a graph neural network based on hierarchical clustering and reduction, for efficient text classification.
HCC-3D: Hierarchical Compensatory Compression for 98% 3D Token Reduction in Vision-Language Models
Liheng Zhang (China University of Petroleum (East China)), Weifeng Liu (China University of Petroleum (East China))
CompressionComputational EfficiencyVision Language ModelPoint Cloud
🎯 What it does: Proposed a hierarchical compensation compression framework named HCC-3D, which significantly reduces the computational cost of 3D-VLM and improves inference efficiency by compressing 3D point cloud features into 12 tokens.
HCF: Hierarchical Cascade Framework for Distributed Multi-Stage Image Compression
Junhao Cai (Korea University), Changhee Joo (Korea University)
CompressionAuto EncoderImage
🎯 What it does: Propose a Hierarchical Cascading Framework (HCF) to achieve direct latent space transformation during distributed multi-stage image compression, avoiding pixel domain re-encoding;
HCPO: Hierarchical Conductor-Based Policy Optimization in Multi-Agent Reinforcement Learning
Zejiao Liu (East China University of Science and Technology), Fangfei Li (East China University of Science and Technology)
OptimizationReinforcement LearningBenchmark
🎯 What it does: Propose a hierarchical conductor-driven multi-agent reinforcement learning framework named HCPO, leveraging conductor instructions to enhance the expressiveness of joint policies and achieving centralized training with decentralized execution during training.
HD²-SSC: High-Dimension High-Density Semantic Scene Completion for Autonomous Driving
Zhiwen Yang (Peking University), Yuxin Peng (Peking University)
Autonomous DrivingTransformerPoint Cloud
🎯 What it does: Propose the HD-SSC framework to address the dimension and density gaps in camera-based semantic scene completion, achieving more accurate 3D semantic completion through pixel semantic decoupling and voxel refinement.
HDGS: Hierarchical Dynamic Gaussian Splatting for Urban Driving Scenes
Fudong Ge (State Key Laboratory of Multimodal Artificial Intelligence Systems (MAIS), CASIA), Zhipeng Zhang (Shanghai Jiao Tong University)
Autonomous DrivingGaussian SplattingVideoPoint Cloud
🎯 What it does: Proposed a Hierarchical Dynamic Gaussian Slicing (HDGS) framework that efficiently models city-scale 4D dynamic scenes using an anchor-based Gaussian model and achieves real-time novel view synthesis.
HDRMovieformer: A Transformer Framework and Benchmark for Cinematic SDR-to-HDR Conversion
Xianwei Li (Beijing University of Posts and Telecommunications), Huadong Ma (Beijing University of Posts and Telecommunications)
Image TranslationTransformerVideoBenchmark
🎯 What it does: Propose the HDRMovieformer framework to achieve movie-level SDR-to-HDR conversion, leveraging brightness-guided transformers and color refinement modules to restore high dynamic range images.
Head-Aware KV Cache Compression for Efficient Visual Autoregressive Modeling
Ziran Qin, Weiyao Lin (Tsinghua University)
GenerationCompressionComputational EfficiencyTransformerImage
🎯 What it does: This paper proposes HACK, a training-agnostic head-aware KV cache compression framework, specifically optimized for KV cache accumulation and attention complexity issues in multi-scale generation processes of visual autoregressive (VAR) models.