This paper is on video recognition using Transformers.

C.1 Ablationstudy Ablation study results are provided in Table 3.

In the field of multi-task reinforcement learning, the modular principle, which involves specializing functionalities into different modules and combining them appropriately, has been widely adopted as a promising approach to prevent the negative transfer problem that performance degradation due to conflicts between tasks. However, most of the existing multi-task RL methods only combine shared modules at the task level, ignoring that there may be conflicts within the task. In addition, these methods do not take into account that without constraints, some modules may learn similar functions, resulting in restricting the model's expressiveness and generalization capability of modular methods.In this paper, we propose the Contrastive Modules with Temporal Attention(CMTA) method to address these limitations. CMTA constrains the modules to be different from each other by contrastive learning and combining shared modules at a finer granularity than the task level with temporal attention, alleviating the negative transfer within the task and improving the generalization ability and the performance for multi-task RL.We conducted the experiment on Meta-World, a multi-task RL benchmark containing various robotics manipulation tasks. Experimental results show that CMTA outperforms learning each task individually for the first time and achieves substantial performance improvements over the baselines.

Despite impressive visual fidelity, modern video generative models frequently produce sequences that violate intuitive physical laws, such as objects floating, teleporting, or morphing in ways that defy causality. While humans can easily detect such implausibilities, there remains no robust method for quantitatively assessing physical realism in video. In this work, we explore whether Video-Language Models (VLMs) can be trained to serve as reliable judges of physical plausibility. We find that existing VLMs struggle to identify physics violations, exposing fundamental limitations in their temporal and causal reasoning. To address this, we introduce TRAVL, a fine-tuning recipe that combines a balanced training dataset with a trajectory-aware attention module to improve motion encoding and discrimination in VLMs. To evaluate physical reasoning more rigorously, we propose ImplausiBench, a benchmark of 300 videos (150 real, 150 generated) that removes linguistic biases and isolates visual-temporal understanding. Performance is reported both with gold-standard human judgments and stricter LLM-as-judge metrics. Together, TRAVL and ImplausiBench offer a unified framework for probing and improving physical plausibility in multimodal models, shedding light on a challenging and underexplored aspect of visual-temporal understanding.

Current methods largely rely on subjective observation and rating, which introduce variability and bias, lack physiological grounding, and do not generalize across tasks or individuals . While functional near - infrared spectroscopy (fNIRS) has emerged as a promising non - invasive tool for measuring neural activity during task performance [Nemani et al., 2018; Gao et al., 2021; Keles et al., 2021], prior models -- including our own -- have largely relied on extensive preprocessing pipelines, lacked physiological interpretability, and demonstrated limited generalization across tasks or stress conditions . Recent sub - region - based analyses have shown that localized activation patterns in prefrontal and motor - related areas can distinguish between proficient and non - proficient performers [Nemani et al., 2018, Gao et al., 2021, Subedi et al., 2025]. The complex spatiotemporal dynamics captured in these signals contain rich information about cognitive and motor processes that could enable more robust, generalizable assessments across diverse settings. However, existing models tend to be task - specific and require retraining when applied to new procedures, impeding scalability . These limitations highlight the need for a domain - specific foundation model -- an architecture trained to extract reusable neural representations across various tasks, individuals, and conditions. In contrast to conventional pipelines, such a model would support modular adaptation to new contexts using minimal data, while retaining interpretability. Foundation models, initially developed in natural language processing and vision [ Lu et.

Traffic forecasting is a key task in the field of Intelligent Transportation Systems. Recent research on traffic forecasting has mainly focused on combining graph neural networks (GNNs) with other models. However, GNNs only consider short-range spatial information. In this study, we present a novel model termed LSTAN-GERPE (Lightweight Spatio-Temporal Attention Network with Graph Embedding and Rotational Position Encoding). This model leverages both Temporal and Spatial Attention mechanisms to effectively capture long-range traffic dynamics. Additionally, the optimal frequency for rotational position encoding is determined through a grid search approach in both the spatial and temporal attention mechanisms. This systematic optimization enables the model to effectively capture complex traffic patterns. The model also enhances feature representation by incorporating geographical location maps into the spatio-temporal embeddings. Without extensive feature engineering, the proposed method in this paper achieves advanced accuracy on the real-world traffic forecasting datasets PeMS04 and PeMS08.

In the field of multi-task reinforcement learning, the modular principle, which involves specializing functionalities into different modules and combining them appropriately, has been widely adopted as a promising approach to prevent the negative transfer problem that performance degradation due to conflicts between tasks. However, most of the existing multi-task RL methods only combine shared modules at the task level, ignoring that there may be conflicts within the task. In addition, these methods do not take into account that without constraints, some modules may learn similar functions, resulting in restricting the model's expressiveness and generalization capability of modular methods.In this paper, we propose the Contrastive Modules with Temporal Attention(CMTA) method to address these limitations. CMTA constrains the modules to be different from each other by contrastive learning and combining shared modules at a finer granularity than the task level with temporal attention, alleviating the negative transfer within the task and improving the generalization ability and the performance for multi-task RL.We conducted the experiment on Meta-World, a multi-task RL benchmark containing various robotics manipulation tasks. Experimental results show that CMTA outperforms learning each task individually for the first time and achieves substantial performance improvements over the baselines.

Despite the considerable progress achieved in the long video generation problem, there is still significant room to improve the consistency of the videos, particularly in terms of smoothness and transitions between scenes. We address these issues to enhance the consistency and coherence of videos generated with either single or multiple prompts. We propose the Time-frequency based temporal Attention Reweighting Algorithm (TiARA), which meticulously edits the attention score matrix based on the Discrete Short-Time Fourier Transform. Our method is supported by a theoretical guarantee, the first-of-its-kind for frequency-based methods in diffusion models. For videos generated by multiple prompts, we further investigate key factors affecting prompt interpolation quality and propose PromptBlend, an advanced prompt interpolation pipeline. The efficacy of our proposed method is validated via extensive experimental results, exhibiting consistent and impressive improvements over baseline methods. The code will be released upon acceptance.

Modeling temporal characteristics plays a significant role in the representation learning of audio waveform. We propose Contrastive Long-form Language-Audio Pretraining (\textbf{CoLLAP}) to significantly extend the perception window for both the input audio (up to 5 minutes) and the language descriptions (exceeding 250 words), while enabling contrastive learning across modalities and temporal dynamics. Leveraging recent Music-LLMs to generate long-form music captions for full-length songs, augmented with musical temporal structures, we collect 51.3K audio-text pairs derived from the large-scale AudioSet training dataset, where the average audio length reaches 288 seconds. We propose a novel contrastive learning architecture that fuses language representations with structured audio representations by segmenting each song into clips and extracting their embeddings. With an attention mechanism, we capture multimodal temporal correlations, allowing the model to automatically weigh and enhance the final fusion score for improved contrastive alignment. Finally, we develop two variants of the CoLLAP model with different types of backbone language models. Through comprehensive experiments on multiple long-form music-text retrieval datasets, we demonstrate consistent performance improvement in retrieval accuracy compared with baselines. We also show the pretrained CoLLAP models can be transferred to various music information retrieval tasks, with heterogeneous long-form multimodal contexts.