Submission Instructions
Template: Use the PKU-IAI Technical Report (TR) Template available on the LaTeX Template page. Do not use the Essay Template; it is different from the TR template.
Oral and Written Presentation Rubrics
- Soundness
- Logical consistency and coherence of arguments
- Accuracy of facts, data, and references
- Valid methodology and sound reasoning
- Credible sources and evidence to support claims
- Absence of logical fallacies or contradictions
- Novelty
- Original ideas or unique perspectives presented
- Innovative approaches/perspective to existing problems
- Fresh insights or unexpected connections
- Creative solutions or methodologies
- Contribution of new knowledge to the field
- Depth
- Comprehensive coverage of the topic
- Thorough analysis and detailed examination
- Complex understanding demonstrated beyond surface level
- Multiple perspectives or dimensions explored
- Substantial research and investigation evident
- Presentation
- Clear and engaging delivery style
- Well-organized structure and logical flow
- Effective use of visual aids
- Appropriate timing and pace
- Strong audience engagement and communication skills
- Technical Details
- Accuracy of technical information and specifications
- Proper use of terminology and concepts
- Clear explanation of methods and processes
- Sufficient detail for understanding and replication
- Appropriate level of technical complexity for audience
Project List
- BreathSync: 基于多模态人工智能的智能呼吸调节与冥想生成系统
- CivRealm: 基于文明演化的多智能体学习与推理决策仿真
- CogSynth: 基于生成式AI的3D物体合成系统及概念学习平台
- DebateTroll: 基于论证挖掘与逻辑谬误检测的智能辩论策略生成系统
- GateRL-Humanoid: 基于门控多头强化学习的人形机器人控制系统
- MemFlowAI：基于多智能体的长文本记忆增强系统
- MetaThrow: 基于元强化学习的机器人投掷系统
- OpenMic: 基于多智能体框架的智能脱口秀生成系统
- PersonalGUI: 个性化智能手机界面交互智能体系统
- PoemCloud: 古诗词计算审美代理
- RealPoem: 基于人类声道物理建模的古音诗歌朗诵系统
- SmallVille-LE: 语言涌现与演化多智能体实验
- SymbolGen: 基于概念图驱动的认知符号生成研究
- ToM-BrainNet：基于神经影像与大模型内部表征的心理理论对齐与机制探索
- TowerCollapse: 基于神经网络的塔倒塌预测模型
- UrbanGovSim: 基于游戏平台的高仿真城市治理模拟
- 基于Neural Operator的汽车空气动力学建模
- 基于Virtual Tool Game实现类人的工具使用学习算法
- 基于因果推断与机器学习的儿童期不良经历（ACE）长期结果机制研究
- 基于多智能体框架的即时元认知调节系统
- 基于多模态大模型的人形机器人动作驱动
- 基于多模态数据的创伤后应激障碍发生风险预测模型
- 基于多模态数据的指导性写作暴露干预匹配模型
- 基于多模态长程数据的呼吸干预效果预测模型
- 基于文本叙事的心理健康水平与韧性预测系统
- 基于童年负性事件与手机使用习惯预测自杀行为
- 基于机器学习的移动机器人运动规划方法