Introduction to LLM agents Couger AI 

Autonomous Agent • Computational systems that sense and act autonomously the complex dynamics in the enviroment, and by doing so realize a set of goals or tasks for which they are designed. 

What are agents for? • Autonomous agents are recognized as a symbol of AGI. • accomplish tasks through self- directed planning and actions. 

Challenges and the LLM solution • Enviroment representations. • World knowledge comprehension. • Complexity of human-like decision making. • Appropriate actions/human-like behaviors selection. 

LLM-driven Agent Architecture 

Agent Profile • The information for the role of the agent. • age • gender • career • Generation method • Handicrafting • LLM generation • Real-world human simulation 

Memory • Type • short-term: dynamics that is important for understanding current state and making decisions. • current conversation history • scene description • long-term: information that are important over time. • past decision-making trajectories • presets of behaviors 

Decision-making • Reasoning • Single-path • Chain of thought • Multi-path • Tree of thoughs • Feedback Planning • Environment feedback • objective in the virtual world • Human feedback • human supervision 

Action • Action space • Leveraging knowledge • plan • converse • Using external tools • call APIs • search knowledge base 

Example - Recommendation RecMind 

Recommendation • Consider both customer information and common information • Decompose tasks and select the best reasoning path • Execute tasks with approprite actions 

Example - World Simulation Generative Agents 

World Simulation • Streaming memory 

• Make the plan • Conversation • Idea spreading 14/33 Emergent Communications 

Example - Role playing 

Example - Self emotion simulation • Responses are influenced by self-emotions. 16/33 

Agents and self-emotion simulation[6] • Creation of schedules based on profiles. • Generation of random events. • Analysis of self-emotions. • Communication with self- emotions. 17/33 [6]Zhang, Q., Naradowsky, J., & Miyao, Y. (2024). Self-Emotion Blended Dialogue Generation in Social Simulation Agents. arXiv preprint arXiv:2408.01633. 

Analysis • Positive self-emotions lead to more optimistic startegies and negative self-emotions lead to more pessimistic ones. 18/33 

Self emotion - group discussion • The self-emotion of a member could influence the disussion results of a group. 19/33 

Analysis • Agent with self-emotion incorporated behaves differently. 20/33 

Analysis • Agents with negative self-emotions express more personal opinions, leading to more disagreements. • Positive self-emotions lead to quick agreements and focus on details. 21/33 

A typical workflow in real world • Human supervision can be involved in the interaction loop to improve the performance. 

Conclusion • LLM agents are good approach to • real-world problem representation and simulation • tasks with complicated procedures and external knowledge • applications that involves self-learning and self-improvement