Large language models (LLMs) struggle on processing complicated observations in interactive decision making. To alleviate this issue, we propose a simple hierarchical prompting approach. Diverging from previous prompting approaches that always put the \emph{full} observation~(\eg a web page) to the prompt, we propose to first construct an action-aware observation which is more \emph{condensed} and \emph{relevant} with a dedicated \summ prompt. The \actor prompt then predicts the next action based on the summarized history. While our method has broad applicability, we particularly demonstrate its efficacy in the complex domain of web navigation where a full observation often contains redundant and irrelevant information. Our approach outperforms the previous state-of-the-art prompting mechanism with the same LLM by 6.2\% on task success rate, demonstrating its potential on interactive decision making tasks with long observation traces.

翻译：大语言模型（LLMs）在处理交互式决策中的复杂观察信息时面临挑战。为缓解这一问题，我们提出了一种简单的分层提示方法。与以往总是将完整观察（如网页）直接置于提示中的方法不同，我们首先通过专门的摘要提示构建更精简、更相关的操作感知观察信息。随后，行动提示基于摘要历史信息预测下一步动作。尽管我们的方法具有广泛适用性，我们特别在网页导航这一复杂领域验证了其有效性——该场景下完整观察常包含冗余和无关信息。在相同大语言模型下，我们的方法相比先前最先进的提示机制，任务成功率提升了6.2%，展现了其在长观察序列的交互式决策任务中的潜力。