Malicious content generated by large language models (LLMs) can pose varying degrees of harm. Although existing LLM-based moderators can detect harmful content, they struggle to assess risk levels and may miss lower-risk outputs. Accurate risk assessment allows platforms with different safety thresholds to tailor content filtering and rejection. In this paper, we introduce per-topic severity rubrics for 11 harmful topics and build BingoGuard, an LLM-based moderation system designed to predict both binary safety labels and severity levels. To address the lack of annotations on levels of severity, we propose a scalable generate-then-filter framework that first generates responses across different severity levels and then filters out low-quality responses. Using this framework, we create BingoGuardTrain, a training dataset with 54,897 examples covering a variety of topics, response severity, styles, and BingoGuardTest, a test set with 988 examples explicitly labeled based on our severity rubrics that enables fine-grained analysis on model behaviors on different severity levels. Our BingoGuard-8B, trained on BingoGuardTrain, achieves the state-of-the-art performance on several moderation benchmarks, including WildGuardTest and HarmBench, as well as BingoGuardTest, outperforming best public models, WildGuard, by 4.3\%. Our analysis demonstrates that incorporating severity levels into training significantly enhances detection performance and enables the model to effectively gauge the severity of harmful responses.

Commonsense reasoning (CR) has been studied in many pieces of domain and has achieved great progress with the aid of large datasets. Unfortunately, most existing CR datasets are built in English, so most previous work focus on English. Furthermore, as the annotation of commonsense reasoning is costly, it is impossible to build a large dataset for every novel task. Therefore, there are growing appeals for Cross-lingual Low-Resource Commonsense Reasoning, which aims to leverage diverse existed English datasets to help the model adapt to new cross-lingual target datasets with limited labeled data. In this paper, we propose a multi-source adapter for cross-lingual low-resource Commonsense Reasoning (MetaXCR). In this framework, we first extend meta learning by incorporating multiple training datasets to learn a generalized task adapters across different tasks. Then, we further introduce a reinforcement-based sampling strategy to help the model sample the source task that is the most helpful to the target task. Finally, we introduce two types of cross-lingual meta-adaption methods to enhance the performance of models on target languages. Extensive experiments demonstrate MetaXCR is superior over state-of-the-arts, while being trained with fewer parameters than other work.

Analyzing student behavior in educational scenarios is crucial for enhancing teaching quality and student engagement. Existing AI-based models often rely on classroom video footage to identify and analyze student behavior. While these video-based methods can partially capture and analyze student actions, they struggle to accurately track each student's actions in physical education classes, which take place in outdoor, open spaces with diverse activities, and are challenging to generalize to the specialized technical movements involved in these settings. Furthermore, current methods typically lack the ability to integrate specialized pedagogical knowledge, limiting their ability to provide in-depth insights into student behavior and offer feedback for optimizing instructional design. To address these limitations, we propose a unified end-to-end framework that leverages human activity recognition technologies based on motion signals, combined with advanced large language models, to conduct more detailed analyses and feedback of student behavior in physical education classes. Our framework begins with the teacher's instructional designs and the motion signals from students during physical education sessions, ultimately generating automated reports with teaching insights and suggestions for improving both learning and class instructions. This solution provides a motion signal-based approach for analyzing student behavior and optimizing instructional design tailored to physical education classes. Experimental results demonstrate that our framework can accurately identify student behaviors and produce meaningful pedagogical insights.

Most current federated learning frameworks are modeled as static processes, ignoring the dynamic characteristics of the learning system. Under the limited communication budget of the central server, the flexible model architecture of a large number of clients participating in knowledge transfer requires a lower participation rate, active clients have uneven contributions, and the client scale seriously hinders the performance of FL. We consider a more general and practical federation scenario and propose a system heterogeneous federation method based on data-free knowledge distillation and two-way contrast (HFedCKD). We apply the Inverse Probability Weighted Distillation (IPWD) strategy to the data-free knowledge transfer framework. The generator completes the data features of the nonparticipating clients. IPWD implements a dynamic evaluation of the prediction contribution of each client under different data distributions. Based on the antibiased weighting of its prediction loss, the weight distribution of each client is effectively adjusted to fairly integrate the knowledge of participating clients. At the same time, the local model is split into a feature extractor and a classifier. Through differential contrast learning, the feature extractor is aligned with the global model in the feature space, while the classifier maintains personalized decision-making capabilities. HFedCKD effectively alleviates the knowledge offset caused by a low participation rate under data-free knowledge distillation and improves the performance and stability of the model. We conduct extensive experiments on image and IoT datasets to comprehensively evaluate and verify the generalization and robustness of the proposed HFedCKD framework.

Accurate sign language understanding serves as a crucial communication channel for individuals with disabilities. Current sign language translation algorithms predominantly rely on RGB frames, which may be limited by fixed frame rates, variable lighting conditions, and motion blur caused by rapid hand movements. Inspired by the recent successful application of event cameras in other fields, we propose to leverage event streams to assist RGB cameras in capturing gesture data, addressing the various challenges mentioned above. Specifically, we first collect a large-scale RGB-Event sign language translation dataset using the DVS346 camera, termed VECSL, which contains 15,676 RGB-Event samples, 15,191 glosses, and covers 2,568 Chinese characters. These samples were gathered across a diverse range of indoor and outdoor environments, capturing multiple viewing angles, varying light intensities, and different camera motions. Due to the absence of benchmark algorithms for comparison in this new task, we retrained and evaluated multiple state-of-the-art SLT algorithms, and believe that this benchmark can effectively support subsequent related research. Additionally, we propose a novel RGB-Event sign language translation framework (i.e., M$^2$-SLT) that incorporates fine-grained micro-sign and coarse-grained macro-sign retrieval, achieving state-of-the-art results on the proposed dataset. Both the source code and dataset will be released on https://github.com/Event-AHU/OpenESL.

This study proposes a new deep learning method for reconstructing depth images of moving objects within a specific area using Wi-Fi channel state information (CSI). The Wi-Fi-based depth imaging technique has novel applications in domains such as security and elder care. However, reconstructing depth images from CSI is challenging because learning the mapping function between CSI and depth images, both of which are high-dimensional data, is particularly difficult. To address the challenge, we propose a new approach called Wi-Depth. The main idea behind the design of Wi-Depth is that a depth image of a moving object can be decomposed into three core components: the shape, depth, and position of the target. Therefore, in the depth-image reconstruction task, Wi-Depth simultaneously estimates the three core pieces of information as auxiliary tasks in our proposed VAE-based teacher-student architecture, enabling it to output images with the consistency of a correct shape, depth, and position. In addition, the design of Wi-Depth is based on our idea that this decomposition efficiently takes advantage of the fact that shape, depth, and position relate to primitive information inferred from CSI such as angle-of-arrival, time-of-flight, and Doppler frequency shift.

Continual Reinforcement Learning (CRL) is essential for developing agents that can learn, adapt, and accumulate knowledge over time. However, a fundamental challenge persists as agents must strike a delicate balance between plasticity, which enables rapid skill acquisition, and stability, which ensures long-term knowledge retention while preventing catastrophic forgetting. In this paper, we introduce SSDE, a novel structure-based approach that enhances plasticity through a fine-grained allocation strategy with Structured Sparsity and Dormant-guided Exploration. SSDE decomposes the parameter space into forward-transfer (frozen) parameters and task-specific (trainable) parameters. Crucially, these parameters are allocated by an efficient co-allocation scheme under sparse coding, ensuring sufficient trainable capacity for new tasks while promoting efficient forward transfer through frozen parameters. However, structure-based methods often suffer from rigidity due to the accumulation of non-trainable parameters, limiting exploration and adaptability. To address this, we further introduce a sensitivity-guided neuron reactivation mechanism that systematically identifies and resets dormant neurons, which exhibit minimal influence in the sparse policy network during inference. This approach effectively enhance exploration while preserving structural efficiency. Extensive experiments on the CW10-v1 Continual World benchmark demonstrate that SSDE achieves state-of-the-art performance, reaching a success rate of 95%, surpassing prior methods significantly in both plasticity and stability trade-offs (code is available at: https://github.com/chengqiArchy/SSDE).

As machine learning systems become increasingly embedded in human society, their impact on the natural world continues to escalate. Technical evaluations have addressed a variety of potential harms from large language models (LLMs) towards humans and the environment, but there is little empirical work regarding harms towards nonhuman animals. Following the growing recognition of animal protection in regulatory and ethical AI frameworks, we present the Animal Harm Assessment (AHA), a novel evaluation of risks of animal harm in LLM-generated text. Our dataset comprises 1,850 curated questions from Reddit post titles and 2,500 synthetic questions based on 50 animal categories (e.g., cats, reptiles) and 50 ethical scenarios, with further 70-30 public-private split. Scenarios include open-ended questions about how to treat animals, practical scenarios with potential animal harm, and willingness-to-pay measures for the prevention of animal harm. Using the LLM-as-a-judge framework, answers are evaluated for their potential to increase or decrease harm, and evaluations are debiased for the tendency to judge their own outputs more favorably. We show that AHA produces meaningful evaluation results when applied to frontier LLMs, revealing significant differences between models, animal categories, scenarios, and subreddits. We conclude with future directions for technical research and the challenges of building evaluations on complex social and moral topics.

With the rapid advancement of digitalization, various document images are being applied more extensively in production and daily life, and there is an increasingly urgent need for fast and accurate parsing of the content in document images. Therefore, this report presents PP-DocBee, a novel multimodal large language model designed for end-to-end document image understanding. First, we develop a data synthesis strategy tailored to document scenarios in which we build a diverse dataset to improve the model generalization. Then, we apply a few training techniques, including dynamic proportional sampling, data preprocessing, and OCR postprocessing strategies. Extensive evaluations demonstrate the superior performance of PP-DocBee, achieving state-of-the-art results on English document understanding benchmarks and even outperforming existing open source and commercial models in Chinese document understanding. The source code and pre-trained models are publicly available at \href{https://github.com/PaddlePaddle/PaddleMIX}{https://github.com/PaddlePaddle/PaddleMIX}.

Sampling from diffusion models involves a slow iterative process that hinders their practical deployment, especially for interactive applications. To accelerate generation speed, recent approaches distill a multi-step diffusion model into a single-step student generator via variational score distillation, which matches the distribution of samples generated by the student to the teacher's distribution. However, these approaches use the reverse Kullback-Leibler (KL) divergence for distribution matching which is known to be mode seeking. In this paper, we generalize the distribution matching approach using a novel $f$-divergence minimization framework, termed $f$-distill, that covers different divergences with different trade-offs in terms of mode coverage and training variance. We derive the gradient of the $f$-divergence between the teacher and student distributions and show that it is expressed as the product of their score differences and a weighting function determined by their density ratio. This weighting function naturally emphasizes samples with higher density in the teacher distribution, when using a less mode-seeking divergence. We observe that the popular variational score distillation approach using the reverse-KL divergence is a special case within our framework. Empirically, we demonstrate that alternative $f$-divergences, such as forward-KL and Jensen-Shannon divergences, outperform the current best variational score distillation methods across image generation tasks. In particular, when using Jensen-Shannon divergence, $f$-distill achieves current state-of-the-art one-step generation performance on ImageNet64 and zero-shot text-to-image generation on MS-COCO. Project page: https://research.nvidia.com/labs/genair/f-distill