Series Ⅲ

Lecture Guests

Shengyao Lu
Assistant Professor @ University of Victoria (UVic)
shengyaolu@uvic.ca

Haipeng Chen
Assistant Professor @ William & Mary, Graduate Program Director of Data Science Department

Schedule

Date	Guest Lecture	Supplemental Readings
November 4th	Towards Graph-Language Systems Shengyao Lu, University of Victoria (UVic) Key Topics: Bridging graph neural networks and large language models Explainable AI for graph-structured knowledge Natural language interaction with structured knowledge Reasoning and decision-making in graph-language systems Applications and challenges in building interpretable AI systems Recording	To be updated
December 19th	Turing Smith Machine: AI of All, by All, for All: Decentralized AI aka AI + Blockchain Yuxi Li, University of Alberta, AI4All Institute First principles: Learning from experience Iterative improvement based on ground truth Exploration, exploitation, and evolution Research: Turing Smith Machine: decentralized AI platform integrating AI and blockchain Modules for infrastructure, decision making, marketplace, and applications Cross-disciplinary approach: AI, computer science, economics, behavioral sciences Building the "AI flywheel" through learning algorithms and incentive mechanisms Applications: finance x AI x blockchain, software real world assets (RWAs), decentralized education and research Recording	To be updated
January 14th	Reinforcement Learning for Language Models Haipeng Chen, William & Mary Key Topics: RL for fine-tuning the LLM: Reinforcement Learning from Human Feedback (RLHF) to align model behavior with human intent Text generation as a sequential decision process: refining pretrained models into agents Optimizing for quality, helpfulness, and safety rather than just raw likelihood RL for training separate, auxiliary models: smaller-scale models that interact with or guide LLMs Adaptive feedback loops: learning to evaluate, interpret, or modulate LLM outputs Extending LLM performance in dynamic or multi-agent environments Recording	To be updated

Date

Guest Lecture

Supplemental Readings

November 4th

Towards Graph-Language Systems
Shengyao Lu, University of Victoria (UVic)

Key Topics:

Bridging graph neural networks and large language models
Explainable AI for graph-structured knowledge
Natural language interaction with structured knowledge
Reasoning and decision-making in graph-language systems
Applications and challenges in building interpretable AI systems

Recording

To be updated

December 19th

Turing Smith Machine: AI of All, by All, for All: Decentralized AI aka AI + Blockchain
Yuxi Li, University of Alberta, AI4All Institute

First principles:

Learning from experience
Iterative improvement based on ground truth
Exploration, exploitation, and evolution

Research:

Turing Smith Machine: decentralized AI platform integrating AI and blockchain
Modules for infrastructure, decision making, marketplace, and applications
Cross-disciplinary approach: AI, computer science, economics, behavioral sciences
Building the "AI flywheel" through learning algorithms and incentive mechanisms
Applications: finance x AI x blockchain, software real world assets (RWAs), decentralized education and research

Recording

To be updated

January 14th

Reinforcement Learning for Language Models
Haipeng Chen, William & Mary

Key Topics:

RL for fine-tuning the LLM: Reinforcement Learning from Human Feedback (RLHF) to align model behavior with human intent
Text generation as a sequential decision process: refining pretrained models into agents
Optimizing for quality, helpfulness, and safety rather than just raw likelihood
RL for training separate, auxiliary models: smaller-scale models that interact with or guide LLMs
Adaptive feedback loops: learning to evaluate, interpret, or modulate LLM outputs
Extending LLM performance in dynamic or multi-agent environments

Recording

To be updated