Human-defined creativity is highly abstract, posing a challenge for multimodal large language models (MLLMs) to comprehend and assess creativity that aligns with human judgments. The absence of an existing benchmark further exacerbates this dilemma. To this end, we propose CreBench, which consists of two key components: 1) an evaluation benchmark covering the multiple dimensions from creative idea to process to products; 2) CreMIT (Creativity Multimodal Instruction Tuning dataset), a multimodal creativity evaluation dataset, consisting of 2.2K diverse-sourced multimodal data, 79.2K human feedbacks and 4.7M multityped instructions. Specifically, to ensure MLLMs can handle diverse creativity-related queries, we prompt GPT to refine the human feedback to activate stronger creativity assessment capabilities. CreBench serves as a foundation for building MLLMs that understand human-aligned creativity. Based on the CreBench, we fine-tune open-source general MLLMs, resulting in CreExpert, a multimodal creativity evaluation expert model. Extensive experiments demonstrate that the proposed CreExpert models achieve significantly better alignment with human creativity evaluation compared to state-ofthe-art MLLMs, including the most advanced GPT-4V and Gemini-Pro-Vision.

Cross-modal retrieval maps data under different modalities via semantic relevance. Existing approaches implicitly assume that data pairs are well-aligned and ignore the widely existing annotation noise, i.e., noisy correspondence (NC). Consequently, it inevitably causes performance degradation. Despite attempts that employ the co-teaching paradigm with identical architectures to provide distinct data perspectives, the differences between these architectures primarily stem from random initialization. Thus, the model becomes increasingly homogeneous along with the training process. Consequently, the additional information brought by this paradigm is severely limited. In order to resolve this problem, we introduce Tripartite Learning with Semantic Variation Consistency (TSVC) for robust image-text retrieval. We design a tripartite cooperative learning mechanism comprising a Coordinator, a Master, and an Assistant model. The Coordinator distributes data, and the Assistant model supports the Master model's noisy label prediction with diverse data. Moreover, we introduce a soft label estimation method based on mutual information variation, which quantifies the noise in new samples and assigns corresponding soft labels. We also present a new loss function to enhance robustness and optimize training effectiveness. Extensive experiments on three widely used datasets demonstrate that, even at increasing noise ratios, TSVC exhibits significant advantages in retrieval accuracy and maintains stable training performance.

Few-shot defect multi-classification (FSDMC) is an emerging trend in quality control within industrial manufacturing. However, current FSDMC research often lacks generalizability due to its focus on specific datasets. Additionally, defect classification heavily relies on contextual information within images, and existing methods fall short of effectively extracting this information. To address these challenges, we propose a general FSDMC framework called MVREC, which offers two primary advantages: (1) MVREC extracts general features for defect instances by incorporating the pre-trained AlphaCLIP model. (2) It utilizes a region-context framework to enhance defect features by leveraging mask region input and multi-view context augmentation. Furthermore, Few-shot Zip-Adapter(-F) classifiers within the model are introduced to cache the visual features of the support set and perform few-shot classification. We also introduce MVTec-FS, a new FSDMC benchmark based on MVTec AD, which includes 1228 defect images with instance-level mask annotations and 46 defect types. Extensive experiments conducted on MVTec-FS and four additional datasets demonstrate its effectiveness in general defect classification and its ability to incorporate contextual information to improve classification performance.

Session-Based Recommendation (SBR) based on Graph Neural Networks (GNN) has become a new paradigm for recommender systems, and plays a fundamental role in e-commerce and other relevant domains. Existing graph aggregation methods primarily form node representations by capturing basic relationships between neighboring and central nodes. Despite their encouraging results, the global relationships of items and user intentions within sessions typically change over time, which degrades the effectiveness of existing embedding schemes. To resolve this challenge, we propose a Long and Short-Term Temporal Graph Neural Network (LS-TGNN) for SBR. LS-TGNN employs a novel temporal session graph to aggregate neighborhood information, and models user interests from both long and short-term perspectives. Specifically, we design long-term and short-term encoders to model the long and short-term interests of users, respectively. In order to better model the interests of users in different time dimensions, we introduce an item-granularity method that distinguishes between long and short-term interests. Extensive experiments on three widely used datasets demonstrate that LS-TGNN outperforms existing methods with a large margin.

Network load balancing is important in datacenter network for LLM training with low entropy traffic. Through in-depth experimental investigations, we find that traditional flowlet-based switch rerouting schemes face substantial challenges in achieving an optimal balance between flowlet timeout settings and packet reordering issues under LLM traffic. Thus, we present CoLB, an application layer and network layer codesign load balancing scheme. CoLB utilizes size-based flow block instead of a timeout value-based flowlet as the rerouting unit for LLM traffic, where the size of the flow block is adaptive to the collective communication patterns generated by the application workload layer. Besides, CoLB is aware of the link utilization affected by PFC mechanism in the underlying RDMA network and guarantees a more accurate perception of link congestion conditions. We evaluate CoLB with the NS-3 simulator under the collective communication traffic and realistic workloads of LLMs. The simulation results show that CoLB is superior to previous switch-based load balancing schemes. For example, CoLB achieves up to 79.88% lower 99th Flow Completion Time (FCT) slowdown than LetFLow and up to 38.83% than Conweave.