Developing agents capable of fluid gameplay in first/third-person games without API access remains a critical challenge in Artificial General Intelligence (AGI). Recent efforts leverage Vision Language Models (VLMs) as direct controllers, frequently pausing the game to analyze screens and plan action through language reasoning. However, this inefficient paradigm fundamentally restricts agents to basic and non-fluent interactions: relying on isolated VLM reasoning for each action makes it impossible to handle tasks requiring high reactivity (e.g., FPS shooting) or dynamic adaptability (e.g., ACT combat). To handle this, we propose a paradigm shift in gameplay agent design: instead of direct control, VLM serves as a developer, creating specialized execution modules tailored for tasks like shooting and combat. These modules handle real-time game interactions, elevating VLM to a high-level developer. Building upon this paradigm, we introduce GameSense, a gameplay agent framework where VLM develops task-specific game sense modules by observing task execution and leveraging vision tools and neural network training pipelines. These modules encapsulate action-feedback logic, ranging from direct action rules to neural network-based decisions. Experiments demonstrate that our framework is the first to achieve fluent gameplay in diverse genres, including ACT, FPS, and Flappy Bird, setting a new benchmark for game-playing agents.

An extensive high-quality instruction dataset is crucial for the instruction tuning process of Large Language Models (LLMs).Recent instruction expansion methods have demonstrated their capability to improve the quality and quantity of existing datasets, by prompting high-performance LLM to generate multiple new instructions from the original ones.However, existing methods focus on constructing multi-perspective prompts (e.g., increasing complexity or difficulty) to expand instructions, overlooking the "Fixed Thinking Pattern" issue of LLMs.This issue arises when repeatedly using the same set of prompts, causing LLMs to rely on a limited set of certain expressions to expand all instructions, potentially compromising the diversity of the final expanded dataset.This paper theoretically analyzes the causes of the "Fixed Thinking Pattern", and corroborates this phenomenon through multi-faceted empirical research.Furthermore, we propose a novel method based on dynamic prompt updating: Global Eye.Specifically, after a fixed number of instruction expansions, we analyze the statistical characteristics of newly generated instructions and then update the prompts.Experimental results show that our method enables LLaMA3-8B and LLaMA2-13B to surpass the performance of open-source LLMs and GPT3.5 across various metrics.

Classification of encrypted traffic is essential to the security of collaborative systems. Traffic pre-trained models (TPTMs) have shown promising results on this task. Existing methods of TPTMs for encrypted traffic classification typically fine-tune the parameters of TPTM to adapt to different downstream tasks. However, this approach requires updating all parameters of the large-scale TPTM during the training stage and storing an entire new model for every task, which consumes significant computational and storage resources. This paper proposes the Efficient Adapter of Traffic (EAT), which improves the parameter efficiency of TPTMs. Specifically, it fixes the original parameters of the TPTM and inserts a learnable module between each layer of the TPTM. In the process of adapting to all downstream tasks, only the learnable module needs to be trained and stored, which makes our method more efficient in terms of computation and storage. To validate the effectiveness of our approach, we conducted experiments on various encrypted traffic datasets. The results demonstrate that our method achieves similar performance to the fine-tuning method by only adjusting 3.4% of the TPTM parameter quantity, which significantly reduces computational and storage resources.

Knowledge Graphs (KGs) often suffer from incomplete knowledge, which restricts their utility.Recently, Contrastive Learning (CL) has been introduced to Knowledge Graph Completion (KGC), significantly improving the discriminative capabilities of KGC models and setting new benchmarks in performance.However, existing contrastive methods primarily focus on individual triples, overlooking the broader structural connectivities of KGs.This narrow focus hampers a more comprehensive understanding of the graph's structural knowledge.To address this gap, we propose StructKGC, a novel contrastive learning framework designed to flexibly accommodate the diverse topologies inherent in KGs.We introduce four contrastive tasks tailored to KG data: Vertex-level CL, Neighbor-level CL, Path-level CL, and Relation composition level CL.These tasks are trained synergistically during the fine-tuning of pretrained language models (PLMs), allowing for a more nuanced capture of subgraph semantics.To validate the effectiveness of our method, we perform a comprehensive set of experiments on several real-world datasets.The experimental results demonstrate that our approach achieves SOTA performance under standard supervised and low-resource settings.Furthermore, we observe that the various structure-aware tasks introduced can mutually reinforce each other, resulting in consistent performance enhancements.

Mixing data augmentation methods have been widely used in text classification recently. However, existing methods do not control the quality of augmented data and have low model explainability. To tackle these issues, this paper proposes an explainable text classification solution based on attentive and targeted mixing data augmentation, ATMIX. Instead of selecting data for augmentation without control, ATMIX focuses on the misclassified training samples as the target for augmentation to better improve the model's capability. Meanwhile, to generate meaningful augmented samples, it adopts a self-attention mechanism to understand the importance of the subsentences in a text, and cut and mix the subsentences between the misclassified and correctly classified samples wisely. Furthermore, it employs a novel dynamic augmented data selection framework based on the loss function gradient to dynamically optimize the augmented samples for model training. In the end, we develop a new model explainability evaluation method based on subsentence attention and conduct extensive evaluations over multiple real-world text datasets. The results indicate that ATMIX is more effective with higher explainability than the typical classification models, hidden-level, and input-level mixup models.