| 1 |
Who Will Top the Charts? Multimodal Music Popularity Prediction via Adaptive Fusion of Modality Experts and Temporal Engagement Modeling |
提出GAMENet，通过自适应融合多模态专家和时序建模预测音乐流行度。 |
large language model multimodal |
|
|
| 2 |
AgenticCyber: A GenAI-Powered Multi-Agent System for Multimodal Threat Detection and Adaptive Response in Cybersecurity |
AgenticCyber：基于生成式AI的多智能体系统，用于网络安全中的多模态威胁检测与自适应响应 |
multimodal |
|
|
| 3 |
Echo-CoPilot: A Multi-View, Multi-Task Agent for Echocardiography Interpretation and Reporting |
Echo-CoPilot：用于心动超声解读和报告的多视角多任务智能体 |
large language model foundation model |
|
|
| 4 |
DaGRPO: Rectifying Gradient Conflict in Reasoning via Distinctiveness-Aware Group Relative Policy Optimization |
DaGRPO：通过区分性感知的组相对策略优化来纠正推理中的梯度冲突 |
large language model instruction following |
|
|
| 5 |
Less Is More for Multi-Step Logical Reasoning of LLM Generalisation Under Rule Removal, Paraphrasing, and Compression |
提出逻辑推理评估框架，揭示LLM在规则扰动下的泛化能力瓶颈 |
large language model |
|
|
| 6 |
Vec-LUT: Vector Table Lookup for Parallel Ultra-Low-Bit LLM Inference on Edge Devices |
提出Vec-LUT，解决边缘设备上超低比特LLM并行推理的内存带宽瓶颈。 |
large language model |
✅ |
|
| 7 |
GENIUS: An Agentic AI Framework for Autonomous Design and Execution of Simulation Protocols |
GENIUS：一个用于自主设计和执行模拟协议的Agentic AI框架 |
large language model |
|
|
| 8 |
Protecting Bystander Privacy via Selective Hearing in Audio LLMs |
提出SH-Bench和BPFT，提升音频LLM在多说话人场景下的旁观者隐私保护能力。 |
large language model |
|
|
| 9 |
DUET: Agentic Design Understanding via Experimentation and Testing |
DUET：通过实验和测试实现Agentic设计理解，提升硬件设计任务性能 |
large language model |
|
|