AAAI 2026 论文列表

Multimodal Domain Generalization (MMDG) leverages the complementary strengths of multiple modalities to enhance model generalization on unseen domains. A central challenge in multimodal learning is optimization imbalance, where modalities converge at different speeds during training. This imbalance leads to unequal gradient contributions, allowing some modalities to dominate the learning process while others lag behind. Existing balancing strategies typically regulate each modality’s gradient contribution based on its classification performance on the source domain to alleviate this issue. However, relying solely on source-domain accuracy neglects a key insight in MMDG: modalities that excel on the source domain may generalize poorly to unseen domains, limiting cross-domain gains. To overcome this limitation, we propose Gradient Modulation Projection (GMP), a unified strategy that promotes balanced optimization in MMDG. GMP first decouples gradients associated with classification and domain-invariance objectives. It then modulates each modality’s gradient based on semantic and domain confidence. Moreover, GMP dynamically adjusts gradient projections by tracking the relative strength of each task, mitigating conflicts between classification and domain-invariant learning within modality-specific encoders. Extensive experiments demonstrate that GMP achieves state-of-the-art performance and integrates flexibly with diverse MMDG methods, significantly improving generalization across multiple benchmarks.

Negative sampling (NS) strategies play a crucial role in knowledge graph representation. In order to overcome the limitations of existing negative sampling strategies, such as vulnerability to false negatives, limited generalization, and lack of control over sample hardness, we propose DANS-KGC (Diffusion-based Adaptive Negative Sampling for Knowledge Graph Completion). DANS-KGC comprises three key components: the Difficulty Assessment Module (DAM), the Adaptive Negative Sampling Module (ANS), and the Dynamic Training Mechanism (DTM). DAM evaluates the learning difficulty of entities by integrating semantic and structural features. Based on this assessment, ANS employs a conditional diffusion model with difficulty-aware noise scheduling, leveraging semantic and neighborhood information during the denoising phase to generate negative samples of diverse hardness. DTM further enhances learning by dynamically adjusting the hardness distribution of negative samples throughout training, enabling a curriculum-style progression from easy to hard examples. Extensive experiments on six benchmark datasets demonstrate the effectiveness and generalization ability of DANS-KGC, with the method achieving state-of-the-art results on all three evaluation metrics for the UMLS and YAGO3-10 datasets.

Mixture of Experts (MoE) models have emerged as a promising approach to scale language models efficiently by activating only a subset of parameters for each input. However, deploying these models under GPU memory constraints remains challenging, as existing offloading strategies incur significant overhead from CPU-GPU data transfers. While prior work has explored prefetching techniques to mitigate this bottleneck, these methods require costly fallback mechanisms when predictions fail. Since expert transfers cannot be canceled once initiated, the correct experts need to be loaded on demand sequentially, introducing additional latency. To address this, we present CommitMoE, a novel approach featuring a Commit Router that makes execution decisions based on expert predictions without fallback mechanisms. Our key insight reveals that router certainty strongly correlates with prediction accuracy, while in low-certainty scenarios, the model output demonstrates inherent robustness to expert selection. Leveraging this insight to design a systems-level solution, CommitMoE achieves 1.3× to 9.4× faster inference across different environments and datasets compared to state-of-the-art offloading frameworks while maintaining model quality.

Covariance matrix estimation in high dimensions is a fundamental problem in machine learning and signal processing. A common structural assumption used to mitigate the challenges posed by high dimensionality is sparsity, which posits that most variable pairs exhibit negligible correlations. In this paper, we revisit the classical problem of positive definite sparse covariance estimation (PDSCE) introduced by Rothman (2012). Unlike many earlier approaches, this formulation incorporates a logarithmic barrier, which guarantees that the resulting covariance estimator is positive definite and thereby ensures the well-posedness of the estimation problem. However, the inclusion of the logarithmic barrier also leads to nontrivial optimization difficulties. To overcome these difficulties, we propose a dual proximal gradient method (DPGM) for solving the PDSCE problem. In contrast to existing primal-space approaches, DPGM operates directly in the dual space. This dual perspective provides several key advantages. First, DPGM significantly reduces computational costs, because positive definiteness is preserved automatically and no iterative subproblem solvers are required. Second, compared with primal optimization algorithms, DPGM offers stronger theoretical guarantees, including principled step size selection and improved iteration complexity. Extensive numerical experiments demonstrate that DPGM consistently outperforms existing methods, which confirms its effectiveness and scalability for high-dimensional sparse covariance estimation.

The Schatten-p norm, as a class of structure-inducing norms based on singular values, has been widely used to enhance model low-rankness and representation capability due to its flexibility in structural modeling and favorable mathematical properties. However, its potential in cluster distribution modeling has long been overlooked. Therefore, we explore the potential of maximizing the Schatten-p norm as a regularization strategy specifically designed to achieve balanced clustering. This work is the first to investigate its effectiveness in promoting cluster balance. To be specific, maximizing Schatten-p norm effectively guides the assignment of data points, ensuring a more balanced distribution of samples across clusters. We have conducted an in-depth theoretical analysis and validated its effectiveness through extensive clustering experiments. Experimental results demonstrate that, compared to existing methods, this regularization term significantly improves clustering quality and obtain reasonable clustering.

Graph Neural Networks (GNNs) have been studied from two primary perspectives: spectral, which employs global graph signal filtering and is theoretically more expressive, and spatial, which builds on local neighborhood aggregation and generalizes well across diverse graph structures. While spectral GNNs are expected to perform better in theory, they often underperform in practice compared to spatial models. To better understand this gap, we introduce a novel theoretical framework for converting spectral GNNs into the spatial domain, allowing for more intuitive analysis. This transformation reveals that signal looping and repeated high-order aggregation are major causes of over-smoothing in spectral GNNs. By addressing these issues in the spatial domain and converting the model back to the spectral domain, we propose DeloopSGNN, a spectral GNN with improved expressive capacity. Experiments on benchmark datasets show that DeloopSGNN achieves consistently strong performance in terms of accuracy and adversarial robustness, demonstrating that spectral GNNs can benefit significantly from careful architectural design grounded in our proposed framework.

Uncovering causal structures from observational data is crucial for understanding complex systems and making informed decisions. While reinforcement learning (RL) has shown promise in identifying these structures in the form of a directed acyclic graph (DAG), existing methods often lack efficiency, making them unsuitable for online applications. In this paper, we propose MARLIN, an efficient multi-agent RL-based approach for incremental DAG learning. MARLIN uses a DAG generation policy that maps a continuous real-valued space to the DAG space as an intra-batch strategy, then incorporates two RL agents—state-specific and state-invariant—to uncover causal relationships and integrates these agents into an incremental learning framework. Furthermore, the framework leverages a factored action space to enhance parallelization efficiency. Extensive experiments on synthetic and real datasets demonstrate that MARLIN outperforms state-of-the-art methods in terms of both efficiency and effectiveness.

Membership Inference Attack (MIA) aims to determine whether a specific data sample was included in the training dataset of a target model. Traditional MIA approaches rely on shadow models to mimic target model behavior, but their effectiveness diminishes for Large Language Model (LLM)-based recommendation systems due to the scale and complexity of training data. This paper introduces a novel knowledge distillation-based MIA paradigm tailored for LLM-based recommendation systems. Our method constructs a reference model via distillation, applying distinct strategies for member and non-member data to enhance discriminative capabilities. The paradigm extracts fused features (e.g., confidence, entropy, loss, and hidden layer vectors) from the reference model to train an attack model, overcoming limitations of individual features. Extensive experiments on extended datasets (Last.FM, MovieLens, Book-Crossing, Delicious) and diverse LLMs (T5, GPT-2, LLaMA3) demonstrate that our approach significantly outperforms shadow model-based MIAs and individual-feature baselines. The results show its practicality for privacy attacks in LLM-driven recommender systems.

Accurate modeling of temporal point processes is critical for reliable event forecasting and informed decision-making. While historical event sequences provide a foundation for intensity estimation, existing approaches often neglect external covariates whose lagged effects impact future intensities across multiple temporal granularities. To address this gap, we propose Multi-Granularity Integration of External Covariates for Temporal Point Processes (METP), a framework for incorporating lagged external influences into intensity modeling. METP extracts periodic structures and decomposes external covariate series into multiple temporal granularities. At each granularity, a lag-aware calibration module is introduced to align covariates with event dynamics. Finally, a hierarchical mixture-of-experts strategy is employed to integrate the multi-granular external covariates with historical event embeddings, enabling a representation of the conditional intensity function with enhanced information. Extensive experiments on public and proprietary datasets demonstrate that METP consistently outperforms existing methods in predictive accuracy.

We study streaming data with categorical features where the vocabulary of categorical feature values is changing and can even grow unboundedly over time. Feature hashing is commonly used as a pre-processing step to map these categorical values into a feature space of fixed size before learning their embeddings. While these methods have been developed and evaluated for offline or batch settings, in this paper we consider online settings. We show that deterministic embeddings are sensitive to the arrival order of categories and suffer from forgetting in online learning, leading to performance deterioration. To mitigate this issue, we propose a probabilistic hash embedding (PHE) model that treats hash embeddings as stochastic and applies Bayesian online learning to learn incrementally from data. Based on the structure of PHE, we derive a scalable inference algorithm to learn model parameters and infer/update the posteriors of hash embeddings and other latent variables. Our algorithm (i) can handle an evolving vocabulary of categorical items, (ii) is adaptive to new items without forgetting old items, (iii) is implementable with a bounded set of parameters that does not grow with the number of distinct observed values on the stream, and (iv) is invariant to the item arrival order. Experiments in classification, sequence modeling, and recommendation systems in online learning setups demonstrate the superior performance of PHE while maintaining high memory efficiency (consumes as low as 2→4% memory of a one-hot embedding table).

Recent findings suggest that consecutive layers of neural networks with the ReLU activation function fold the input space during the learning process. While many works hint at this phenomenon, an approach to quantify the folding was only recently proposed by means of a space folding measure based on the Hamming distance in the discrete activation space. We generalize the space folding measure to a wider class of activation functions through the introduction of equivalence classes of input data. We then analyze its mathematical and computational properties. Lastly, we link the folding to geometry of adversarial attacks. We underpin our claims with an experimental evaluation.

Supervised learning with tabular data presents unique challenges, including low data sizes, the absence of structural cues, and heterogeneous features spanning both categorical and continuous domains. Unlike vision and language tasks, where models can exploit inductive biases in the data, tabular data lacks inherent positional structure, hindering the effectiveness of self-attention mechanisms. While recent transformer-based models like TabTransformer, SAINT, and FT-Transformer (which we refer to as 3T) have shown promise on tabular data, they typically operate without leveraging structural cues such as positional encodings (PEs), as no prior structural information is usually available. In this work, we find both theoretically and empirically that structural cues, specifically PEs can be a useful tool to improve generalization performance for tabular transformers. We find that PEs impart the ability to reduce the effective rank (a form of intrinsic dimensionality) of the features, effectively simplifying the task by reducing the dimensionality of the problem, yielding improved generalization. To that end, we propose Tab-PET (PEs for Tabular Transformers), a graph-based framework for estimating and inculcating PEs into embeddings. Inspired by approaches that derive PEs from graph topology, we explore two paradigms for graph estimation: association-based and causality-based. We empirically demonstrate that graph-derived PEs significantly improve performance across 50 classification and regression datasets for 3T. Notably, association-based graphs consistently yield more stable and pronounced gains compared to causality-driven ones. Our work highlights an unexpected role of PEs in tabular transformers, revealing how they can be harnessed to improve generalization.

Efficiently fine-tuning pre-trained models for downstream tasks is a key challenge in the era of foundation models. Parameter-efficient fine-tuning (PEFT) presents a promising solution, achieving performance comparable to full fine-tuning by updating only a small number of adaptation weights per layer. Traditional PEFT methods typically rely on a single expert, where the adaptation weight is a low-rank matrix. However, for complex tasks, the data's inherent diversity poses a significant challenge for such models, as a single adaptation weight cannot adequately capture the features of all samples. To address this limitation, we explore how to integrate multiple small adaptation experts into a compact structure to defeat a large adapter. Specifically, we propose Tucker Adaptation (TuckA), a method with four key properties: (i) We use Tucker decomposition to create a compact 3D tensor where each slice naturally serves as an expert. The low-rank nature of this decomposition ensures that the number of parameters scales efficiently as more experts are added. (ii) We introduce a hierarchical strategy that organizes these experts into groups at different granularities, allowing the model to capture both local and global data patterns. (iii) We develop an efficient batch-level routing mechanism, which reduces the router's parameter size by a factor of L compared to routing at every adapted layer (where L is the number of adapted layers) (iv) We propose data-aware initialization to achieve loss-free expert load balancing based on theoretical analysis. Extensive experiments on benchmarks in natural language understanding, image classification, and mathematical reasoning speak to the efficacy of TuckA, offering a new and effective solution to the PEFT problem.

Federated learning (FL) has enabled training of multilingual large language models (LLMs) on diverse and decentralized multilingual data, especially on low-resource languages. To improve client-specific performance, personalization via the use of parameter-efficient fine-tuning (PEFT) modules such as LoRA is common. This involves a personalization strategy (PS), such as the design of the PEFT adapter structures (e.g., in which layers to add LoRAs and what ranks) and choice of hyperparameters (e.g., learning rates) for fine-tuning. Instead of manual PS configuration, we propose FedP²EFT, a federated learning-to-personalize method for multilingual LLMs in cross-device FL settings. Unlike most existing PEFT structure selection methods, which are prone to overfitting low-data regimes, FedP²EFT collaboratively learns the optimal personalized PEFT structure for each client via Bayesian sparse rank selection. Evaluations on both simulated and real-world multilingual FL benchmarks demonstrate that FedP²EFT largely outperforms existing personalized fine-tuning methods, while complementing other existing FL methods.

Neural Differential Equations (NDEs) excel at modeling continuous-time dynamics, effectively handling challenges such as irregular observations, missing values, and noise. Despite their advantages, NDEs face a fundamental challenge in adopting dropout, a cornerstone of deep learning regularization, making them susceptible to overfitting. To address this research gap, we introduce Continuum Dropout, a universally applicable regularization technique for NDEs built upon the theory of alternating renewal processes. Continuum Dropout formulates the on-off mechanism of dropout as a stochastic process that alternates between active (evolution) and inactive (paused) states in continuous time. This provides a principled approach to prevent overfitting and enhance the generalization capabilities of NDEs. Moreover, Continuum Dropout offers a structured framework to quantify predictive uncertainty via Monte Carlo sampling at test time. Through extensive experiments, we demonstrate that Continuum Dropout outperforms existing regularization methods for NDEs, achieving superior performance on various time series and image classification tasks. It also yields better-calibrated and more trustworthy probability estimates, highlighting its effectiveness for uncertainty-aware modeling.

Communication efficiency in federated learning (FL) remains a critical challenge in resource-constrained environments. While prototype-based FL reduces communication overhead by sharing class prototypes—mean activations in the penultimate layer—instead of model parameters, its efficiency degrades with larger feature dimensions and class counts. We propose TinyProto, which addresses these limitations through Class-wise Prototype Sparsification (CPS) and Adaptive Prototype Scaling (APS). CPS enables structured sparsity by allocating specific dimensions to class prototypes and transmitting only non-zero elements, thereby achieving higher communication efficiency, while APS scales prototypes based on class distributions to improve performance. Our experiments demonstrate that TinyProto reduces communication costs by up to 10x compared to existing methods while improving performance. Beyond communication efficiency, TinyProto offers crucial advantages: it achieves compression without client-side computational overhead and supports heterogeneous architectures, making it particularly suitable for resource-constrained heterogeneous FL scenarios.

Continual instruction tuning (CIT) has emerged as a promising strategy for adapting large language models (LLMs) to new tasks while preserving historical knowledge. Most existing CIT methods have focused on offline CIT (offCIT), which assumes clearly defined task boundaries and allows multiple passes over the data. However, such assumptions rarely hold in real-world scenarios, where data arrive in a streaming fashion and task boundaries are unknown. This setting introduces critical challenges: the absence of task identifiers (task IDs), a significant imbalance in task-specific information, and inaccessibility to previously seen data. In this work, we propose Online Editing with Decoupled Implicit Task (OnEDIT), an online CIT(onCIT) approach to tackle these challenges. OnEDIT leverages a fixed-size adapter for the implicit task, balancing current and past knowledge through editing operations every time step without relying on task IDs or backpropagation. Extensive experiments on CIT benchmarks demonstrate that OnEDIT consistently maintains robust and stable performance, whereas existing state-of-the-art baselines often suffer from performance degradation in online settings. It suggests that OnEDIT achieves superior generalization across diverse task orders and model scales, while maintaining high efficiency and low memory overhead.

We study a matrix completion problem where both the ground truth R matrix and the unknown sampling distribution P over observed entries are low-rank matrices, and share a common subspace. We assume that a large amount M of unlabeled data drawn from the sampling distribution P is available, together with a small amount N of "labeled" data drawn from the same distribution and noisy estimates of the corresponding ground truth entries. This setting is inspired by recommender systems scenarios where the unlabeled data corresponds to "implicit feedback" (consisting in interactions such as purchase, click, etc. ) and the labeled data corresponds to the "explicit feedback", consisting of interactions where the user has given an explicit rating to the item. Leveraging powerful results from the theory of low-rank subspace recovery, together with classic generalization bounds for matrix completion models, we show error bounds consisting of a sum of two error terms corresponding to sample complexities of nd and dr respectively (ignoring log factors), where d is the rank of P and r is the rank of M. In synthetic experiments, we confirm that the true generalization error naturally splits into independent error terms corresponding to the estimations of P and the ground truth matrix G respectively. In real-life experiments on Douban and MovieLens with most explicit ratings removed, we demonstrate that the method can outperform baselines relying only on the explicit ratings, demonstrating that our assumptions provide a valid toy theoretical setting to study the interaction between explicit and implicit feedbacks in recommender systems.

Personalized text-to-image diffusion models have gained increasing attention because they can generate images that contain unique concepts based on limited training data. However, in continual learning scenarios, these models suffer from concept bleed-through, where newly introduced concepts frequently overwrite or interfere with the previously learned concepts. Previous studies have attempted to mitigate this issue at the model adaptation level; however, they failed to fully preserve the distinct semantic representations in the latent space. Thus, this paper proposes an adversarial perturbation-based training strategy to address concept bleed-through in continual learning for personalized diffusion models. The proposed method introduces adversarial perturbations into the training images, which strategically shifts their semantic representations in the latent space to ensure that the newly learned concepts remain distinct and do not interfere with the previously acquired knowledge. Unlike structural modifications to the model, the proposed method operates at the data level, which makes it broadly applicable to existing continual personalization frameworks without increasing model complexity. Experimental results demonstrate that the proposed method significantly improves concept separation while maintaining high image fidelity, offering a solution to enhance the reliability of continual learning in personalized generative models.