ScaffoldAgent: Utility-Guided Dynamic Outline Optimization for Open-Ended Deep Research
Open-ended deep research (OEDR) requires systems to acquire knowledge through multi-round retrieval and generate coherent long-form reports. The outline plays a central role as a structural scaffold that coordinates retrieval, evidence organization, and generation. However, existing methods either fix the outline before writing or refine it with local heuristics, leading to scaffold drift under continuous information accumulation and delayed feedback for evaluating outline modifications. We propose ScaffoldAgent, a utility-guided dynamic outline optimization framework for OEDR. ScaffoldAgent models outline evolution as a structured decision process with three operations: Expansion, Contraction, and Revision, enabling controlled updates to the report scaffold. It further introduces a utility-guided feedback mechanism that estimates the downstream value of each outline operation from retrieval gain, structural coherence, and trial-generation quality. The resulting utility signal guides node selection, operation scheduling, and termination during inference. Experiments on DeepResearch Bench and DeepResearch Gym show that ScaffoldAgent consistently improves long-form report generation and factual grounding over existing deep research agents.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
OPE: Overcoming Information Saturation in Parallel Thinking via Outline-Guided Path Exploration
Parallel thinking has emerged as a new paradigm for large reasoning models (LRMs) in tackling complex problems. Recent methods leverage Reinforcement Learning (RL) to enhance parallel thinking, aiming to address the limi…
Reinforcement LearningMean Field Games for Optimal Investment Under Relative Performance Criteria
In this paper, we study the portfolio optimization problem formulated by Lacker and Soret. They formulate a finite time horizon model that allows agents to be competitive, measuring their utility not only by their absolu…
Portfolio OptimizationSchedule-and-Calibrate: Utility-Guided Multi-Task Reinforcement Learning for Code LLMs
Reinforcement learning (RL) with verifiable rewards has proven effective at post-training LLMs for coding, yet deploying separate task-specific specialists incurs costs that scale with the number of tasks, motivating a u…
Reinforcement LearningVORTEX: Aligning Task Utility and Human Preferences through LLM-Guided Reward Shaping
In social impact optimization, AI decision systems often rely on solvers that optimize well-calibrated mathematical objectives. However, these solvers cannot directly accommodate evolving human preferences, typically exp…
Pap2Pat: Towards Automated Paper-to-Patent Drafting using Chunk-based Outline-guided Generation
The patent domain is gaining attention in natural language processing research, offering practical applications in streamlining the patenting process and providing challenging benchmarks for large language models (LLMs).…