UNO Arena for Evaluating Sequential Decision-Making Capability of Large Language Models
This work addresses the need for better evaluation of sequential decision-making capabilities in LLMs, which is important for AI researchers, but it is incremental as it builds on existing methods like Monte Carlo metrics and reinforcement learning comparisons.
The paper tackles the problem of evaluating sequential decision-making in large language models (LLMs) by proposing the UNO Arena, a benchmark based on the card game UNO, and finds that their TUTRI player, which incorporates reflection and strategy, achieves a notable performance breakthrough compared to vanilla LLM players.
Sequential decision-making refers to algorithms that take into account the dynamics of the environment, where early decisions affect subsequent decisions. With large language models (LLMs) demonstrating powerful capabilities between tasks, we can't help but ask: Can Current LLMs Effectively Make Sequential Decisions? In order to answer this question, we propose the UNO Arena based on the card game UNO to evaluate the sequential decision-making capability of LLMs and explain in detail why we choose UNO. In UNO Arena, We evaluate the sequential decision-making capability of LLMs dynamically with novel metrics based Monte Carlo methods. We set up random players, DQN-based reinforcement learning players, and LLM players (e.g. GPT-4, Gemini-pro) for comparison testing. Furthermore, in order to improve the sequential decision-making capability of LLMs, we propose the TUTRI player, which can involves having LLMs reflect their own actions wtih the summary of game history and the game strategy. Numerous experiments demonstrate that the TUTRI player achieves a notable breakthrough in the performance of sequential decision-making compared to the vanilla LLM player.