用扩散模型生成多样工具组合，解决自动规划中过早锁定选项的问题。
DiG-Plan: Mitigating Early Commitment for Tool-Graph Planning via Diffusion Guidance
- 先用扩散模型生成多种可能的工具组合，再用自回归模型优化依赖关系。
- 在TaskBench上相比传统方法提升10%相对性能，复杂任务提升更明显。
- 适合需要多步推理和工具选择的复杂自动化任务研究者使用。
生成可执行工具计划需从工具库中选取合适子集，这是一个组合搜索问题，解空间呈指数级增长。然而我们发现主流方法存在关键偏差：标准自回归解码存在早期承诺问题，初始令牌选择会严格限制搜索路径。受控实验表明，在相同计算量下，掩码去噪将Pass@10解覆盖率从0.320提升至0.943。为此，我们提出DiG-Plan框架，将组合探索与结构精炼解耦。DiG-Plan采用基于扩散的提议器，通过迭代精炼生成多样化工具集，随后由自回归精炼器完成依赖预测。在TaskBench上，DiG-Plan相较自回归基线提升10%相对性能，复杂组合任务增益最大；API-Bank结果表明该提议-精炼-选择设计在跨领域任务中仍具有效性。代码已开源。
原文摘要 · Abstract (English)
Generating executable tool plans requires selecting appropriate subsets from tool libraries, a combinatorial search problem with an exponentially large solution space. However, we identify a critical misalignment in predominant approaches: standard autoregressive (AR) decoding suffers from early commitment, where initial token choices rigidly constrain the search trajectory. A controlled study shows that masked denoising raises Pass@10 solution coverage from 0.320 to 0.943 over AR sampling under matched compute. Motivated by this, we propose DiG-Plan, a framework that decouples combinatorial exploration from structural refinement. DiG-Plan employs a diffusion-based proposer to generate diverse tool sets via iterative refinement, followed by an AR refiner for dependency prediction. On TaskBench, DiG-Plan improves over AR baselines by a 10% relative margin, with the largest gains on complex compositional tasks; API-Bank results show that the propose-refine-select design remains effective across domains. Code is available at https://github.com/puddingyeah/DiG-Plan.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。