Program Task Description
Category 1: Hard-Design Problems
-
Can you design a tool-based LLM agent to help with a novel domain, such as coding, email monitoring, or food ordering?
- Customer service: sierra.ai
- Regulatory compliance: norm.ai
- Data science: cognition.ai or delphina.ai
- AI scientist: sakana.ai
- Personal assistants: mindy.com
- General agents: hcompany.ai
Category 2: Soft-Design Problems
-
Can you design high-fidelity LLM agents that can simulate humans?
- Develop a simulation (single-agent or multi-agent) of human behavior and draw novel insights from it
- Generative Agents: Interactive Simulacra of Human Behavior
- Can Large Language Models Transform Computational Social Science?
- How do you improve upon the design interface for LLM agents that act as our companions?