Federated Learning (FL) enables collaborative learning from distributed data while preserving the privacy of participating clients. While supervised federated learning with labeled data has made notable strides and achieved success, federated semi-supervised learning (FSSL) lags in its progress. Existing works for FSSL heavily rely on fully-labeled clients, while ignoring the distribution of pseudo-labels generated from skewed unlabeled data. In this work, we offer empirical and theoretical insights into the challenges encountered when applying conventional semi-supervised algorithms in the federated regime. Specifically, we highlight how the inherent data heterogeneity in FSSL can exacerbate issues within the pseudo-labeling process. Motivated by these observations, we propose federated learning with progressive distribution matching (FedPDM) to regularize the distribution of pseudo-labels, aiming to progressively reshape it to align with the ground-truth distribution. The matching problem could be formulated as an optimal transport (OT) problem and efficiently solved by Sinkhorn-Knopp iteration. Through extensive experiments, we demonstrate the superiority of FedPDM on a variety of models and datasets compared with prior arts for FSSL.

In the age of pervasive machine learning applications, protecting digital content from unauthorized use has become a pressing concern. Unlearnable examples (UEs)—data modified with imperceptible perturbations to inhibit model training while preserving human usability—have emerged as a promising approach. However, existing UE methods assume unauthorized trainers have extensive exposure to UEs or that models are trained from scratch, which may not hold in practical scenarios, This paper investigates the effectiveness of UEs under the few-shot learning paradigm, pitching it against prompt learning (PL) models that leverage pretrained vision-language models (VLMs), like CLIP, capable of generalizing to new classes with minimal data. To address this, we introduce an adaptive UE framework to generate unlearnable examples that specifically target the PL process. In addition, we propose a novel UE countermeasure, A3, with cross-modal adversarial feature alignment, specifically designed to circumvent UEs under few-shot PL. Experimental evaluations on 7 datasets show that A3outperforms existing PL methods, achieving up to 33% higher performance in learning from UEs. For example, in the scenario involving ω→-bounded EM perturbations, A3has an average harmonic mean accuracy across 7 datasets of 82.43%, compared to CoCoOp’s baseline of 65.47%. Our findings highlight the limitations of existing UEs against PL and lay the foundation for future data protection mechanisms.

Transformer models face scalability challenges in causal language modeling (CLM) due to inefficient memory allocation for growing keyvalue (KV) caches, which strains compute and storage resources.Existing methods like Grouped Query Attention (GQA) and tokenlevel KV optimization improve efficiency but rely on rigid resource allocation, often discarding "low-priority" tokens or statically grouping them, failing to address the dynamic spectrum of token importance.We propose mixSGA, a novel mixture-of-expert (MoE) approach that dynamically optimizes token-wise computation and memory allocation.Unlike prior approaches, mixSGA retains all tokens while adaptively routing them to specialized experts with varying KV group sizes, balancing granularity and efficiency.Our key novelties include: (1) a token-wise expert-choice routing mechanism guided by learned importance scores, enabling proportional resource allocation without token discard; (2) weight-sharing across grouped attention projections to minimize parameter overhead; and (3) an auxiliary loss to ensure one-hot routing decisions for training-inference consistency in CLMs.Extensive evaluations across Llama3, TinyLlama, OPT, and Gemma2 model families show mixSGA's superiority over static baselines.On instruction-following and continued pretraining tasks, mixSGA achieves higher ROUGE-L and lower perplexity under the same KV budgets.

The increasing emphasis on privacy and data security has driven the adoption of federated learning (FL). Prompt learning (PL), which fine-tunes prompt embeddings of pretrained models, has gained a surge of interest in FL community, marked by the emergence of an influx of federated prompt learning (FPL) algorithms. Despite recent advancements, a systematic understanding of their underlying mechanisms and principled guidelines for deploying these techniques in different FL scenarios remain absent. Moreover, inconsistent experimental protocols, limited evaluation scenarios, and the lack of the proper assessment of centralized PL methods in existing works have obscured the essence of these algorithms. To close these gaps, we introduce a comprehensive benchmark, named F LIP, to achieve standardized FPL evaluation. F LIP assesses the performance of 13 centralized and FPL methods across 3 FL protocols and 12 open datasets, considering 6 distinct evaluation scenarios. Our findings demonstrate that PL maintains strong generalization performance in both in-distribution and out-of-distribution settings with minimal resource consumption, but there is no silver bullet found for diverse FPL scenarios. The results (1) pinpoint the suitable application scenarios of each FPL algorithm, (2) demonstrate the competitiveness of adapted centralized PL methods, and (3) offer notable insights to interpret their effectiveness and remaining challenges. All benchmarks and code are available to facilitate further research in this domain.

Institutions with limited data and computing resources often outsource model training to third-party providers in a semi-honest setting, assuming adherence to prescribed training protocols with pre-defined learning paradigm (e.g., supervised or semi-supervised learning). However, this practice can introduce severe security risks, as adversaries may poison the training data to embed backdoors into the resulting model. Existing detection approaches predominantly rely on statistical analyses, which often fail to maintain universally accurate detection accuracy across different learning paradigms. To address this challenge, we propose a unified backdoor detection framework in the semi-honest setting that exploits cross-examination of model inconsistencies between two independent service providers. Specifically, we integrate central kernel alignment to enable robust feature similarity measurements across different model architectures and learning paradigms, thereby facilitating precise recovery and identification of backdoor triggers. We further introduce backdoor fine-tuning sensitivity analysis to distinguish backdoor triggers from adversarial perturbations, substantially reducing false positives. Extensive experiments demonstrate that our method achieves superior detection performance, improving accuracy by 4.4%, 1.7%, and 10.6% over SoTA baselines across supervised, self-supervised, and autoregressive learning tasks, respectively. Notably, it is the first to effectively detect backdoors in multimodal large language models, further highlighting its broad applicability and advancing secure deep learning.

Data-Free Knowledge Distillation (DFKD) enables knowledge transfer from a pretrained teacher to a light-weighted student without original training data. Existing works are limited by a strong assumption that samples used to pretrain the teacher model are balanced, which is, however, unrealistic for many real-world tasks. In this work, we investigated a pragmatic yet under-explored problem: how to perform DFKD from a teacher model pretrained from imbalanced data. We observe a seemingly counter-intuitive phenomenon, i.e., adversarial DFKD algorithms favour minority classes, while causing a disastrous impact on majority classes. We theoretically prove that a biased teacher could cause severe disparity on different groups of synthetic data in adversarial distillation, which further exacerbates the mode collapse of a generator and consequently degenerates the overall accuracy of a distilled student model. To tackle this problem, we propose a class-adaptive regularization method, aiming to encourage impartial representation learning of a generator among different classes under a constrained learning formulation. We devise a primal-dual algorithm to solve the target optimization problem. Through extensive experiments, we show that our method mitigates the biased learning of majority classes in DFKD and improves the overall performance compared with baselines. Code will be available at https://github.com/ldpbuaa/ipad.

The ability to accurately predict the trajectory of surrounding vehicles is a critical hurdle to overcome on the journey to fully autonomous vehicles. To address this challenge, we pioneer a novel behavior-aware trajectory prediction model (BAT) that incorporates insights and findings from traffic psychology, human behavior, and decision-making. Our model consists of behavior-aware, interaction-aware, priority-aware, and position-aware modules that perceive and understand the underlying interactions and account for uncertainty and variability in prediction, enabling higher-level learning and flexibility without rigid categorization of driving behavior. Importantly, this approach eliminates the need for manual labeling in the training process and addresses the challenges of non-continuous behavior labeling and the selection of appropriate time windows. We evaluate BAT's performance across the Next Generation Simulation (NGSIM), Highway Drone (HighD), Roundabout Drone (RounD), and Macao Connected Autonomous Driving (MoCAD) datasets, showcasing its superiority over prevailing state-of-the-art (SOTA) benchmarks in terms of prediction accuracy and efficiency. Remarkably, even when trained on reduced portions of the training data (25%), our model outperforms most of the baselines, demonstrating its robustness and efficiency in predicting vehicle trajectories, and the potential to reduce the amount of data required to train autonomous vehicles, especially in corner cases. In conclusion, the behavior-aware model represents a significant advancement in the development of autonomous vehicles capable of predicting trajectories with the same level of proficiency as human drivers. The project page is available on our GitHub.

Owing to the non-i.i.d. nature of client data, channel neurons in federated-learned models may specialize to distinct features for different clients. Yet, existing channel-sparse federated learning (FL) algorithms prescribe fixed sparsity strategies for client models, and may thus prevent clients from training channel neurons collaboratively. To minimize the impact of sparsity on FL convergence, we propose Flado to improve the alignment of client model update trajectories by tailoring the sparsities of individual neurons in each client. Empirical results show that while other sparse methods are surprisingly impactful to convergence, Flado can not only attain the highest task accuracies with unlimited budget across a range of datasets, but also significantly reduce the amount of floating-point operations (FLOPs) required for training more than by 10× under the same communications budget, and push the Pareto frontier of communication/computation trade-off notably further than competing FL algorithms.

We seek to improve crowd counting as we perceive limits of currently prevalent density map estimation approach on both prediction accuracy and time efficiency. We leverage multilevel pixelation of density map as it helps improve SNR of training data and therefore, reduce prediction error. To achieve a better model, we introduce multilayer gradient fusion for training a density-aware global count regressor. More specifically, on training stage, a backbone network receives gradients from multiple branches to learn the density information, whereas those branches are to be detached to accelerate inference. By taking advantages of such method, our model improves benchmark results on public datasets and exhibits itself to be a new solution to crowd counting problems in practice.