Beyond Outlining: Heterogeneous Recursive Planning for Adaptive Long-form Writing with Language Models
This addresses the need for more adaptive writing agents for users in creative and technical domains, representing a novel method for a known bottleneck rather than an incremental improvement.
The paper tackled the problem of constrained adaptability in long-form writing agents by proposing WriteHERE, a framework using recursive task decomposition and dynamic integration of retrieval, reasoning, and composition, which outperformed state-of-the-art approaches across all automatic evaluation metrics in fiction writing and technical report generation.
Long-form writing agents require flexible integration and interaction across information retrieval, reasoning, and composition. Current approaches rely on predefined workflows and rigid thinking patterns to generate outlines before writing, resulting in constrained adaptability during writing. In this paper we propose WriteHERE, a general agent framework that achieves human-like adaptive writing through recursive task decomposition and dynamic integration of three fundamental task types: retrieval, reasoning, and composition. Our methodology features: 1) a planning mechanism that interleaves recursive task decomposition and execution, eliminating artificial restrictions on writing workflow; and 2) integration of task types that facilitates heterogeneous task decomposition. Evaluations on both fiction writing and technical report generation show that our method consistently outperforms state-of-the-art approaches across all automatic evaluation metrics, demonstrating the effectiveness and broad applicability of our proposed framework. We have publicly released our code and prompts to facilitate further research.