Speaker: Jeremy Tian
He / him / his
Founder & CPO @Lucidic AI, previously @DRW
Jeremy Tian (BS, MS Stanford CS, AI specialization) is a founder and the CPO at Lucidic AI, where he develops algorithms to efficiently improve complex agentic systems. His work focuses primarily on the psychology of how LLMs learn to distill into training techniques with the goal of optimizing model-harness co-training algorithms. He previously researched deep learning at Stanford and developed machine learning algorithms for trading at DRW.
Session
Model-Harness Co-Training: Data-Driven Optimization Toward Pareto-Optimal Agents
Teams with agents in production are chasing the optimal agent. Usually they pick one lever: tune the harness (the prompts, tools, memory, and workflow around the model), or fine-tune the model, but the two depend on each other.