arXiv cs.CL· Yifan Zhu, Guanting Chen, Bing Wei, Haoran Luo·· 11 小时前AI 评分38
LoopWeaver：面向受限长文本生成的分层反馈驱动优化方法
HiFlow: Hierarchical Feedback-Driven Optimization for Constrained Long-Form Text Generation
AI 导读
LoopWeaver 提出一种两阶段离线偏好优化分层工作流，用于受限长文本生成：约束感知规划生成结构化计划，可行性过滤剔除无效计划，共享确定性奖励为计划级与生成级 DPO 构建偏好对。推理时模型保持前馈，先生成并修订完整计划、过滤后再按当前子计划与叙事上下文分段生成。在四个 LongGenBench 风格结构化场景、多个开源骨干模型上，约束准确率与文本质量均有提升，代码与数据已公开。
正文 · 原文
Abstract:Large Language Models face a multi-objective optimization problem in constrained long-text generation, balancing global structure, local coherence, and constraint satisfaction. We propose LoopWeaver, a hierarchical workflow with two sequential offline preference-optimization stages. Constraint-aware planning produces structured plans, feasibility filtering removes invalid plans, and a shared deterministic reward constructs preference pairs for plan-level and generation-level DPO. At inference time, LoopWeaver remains feed-forward: it generates and revises a complete plan, filters the revised plan, and then generates segments conditioned on the current sub-plan and running narrative. We evaluate the method on four LongGenBench-style structured scenarios across multiple backbones. Results show improvements in both constraint accuracy and text quality across all four open-source backbones. Our software and data are publicly available at this https URL.
| Comments: | Accepted by NeurIPS 2026 main conference |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2603.04996 [cs.CL] |
| (or arXiv:2603.04996v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2603.04996 arXiv-issued DOI via DataCite |
|
| Journal reference: | NeurIPS 2026 |
Submission history
From: Guanting Chen [view email]
[v1]
Thu, 5 Mar 2026 09:41:39 UTC (3,160 KB)
[v2]
Fri, 6 Nov 2026 14:59:04 UTC (5,319 KB)
来源：arXiv cs.CL · arxiv.org