PCG-RAG: Progressive Compositional Grounding for Retrieval-Augmented Generation in Complex Web Question Answering

Linyao Zheng, Xuhang Shi, Zhifang Mao, Chongqi Qu, Xiuquan HouarXiv 2026 · arXiv

TL;DR

PCG-RAG is a novel framework that prevents Retrieval-Augmented Generation (RAG) systems from being misled by partial entity matches during complex web searches. By treating initial findings as revisable hypotheses rather than final answers, it progressively refines entity identification and ensures that every fact retrieved is grounded in a complete chain of evidence.

What Problem Does This Paper Address?

When we ask a complex question like, "Which architecture firm redesigned the London power station that became an art museum opening in 2000?", a standard RAG system might find a page about "Battersea Power Station" and immediately assume it's the correct target. This is premature commitment.

If the system then searches for "Battersea redesign architect," it will find WilkinsonEyre. However, Battersea didn't become a museum in 2000—Bankside Power Station (Tate Modern) did. By latching onto a partial match too early, the system retrieves the wrong facts for the wrong entity.

Current multi-step RAG systems struggle with this "identity-dependent" reasoning. They often fail to connect the clues used to identify an entity (e.g., "museum opened in 2000") with the relations they are trying to find (e.g., "redesign architect").

Figure 1: Two retrieval paths from a partial entity match. In (a), committing to Battersea directs the architect query toward a fact about the wrong subject. In (b), acquiring the missing identifying evidence establishes Bankside, whose conversion is attributed to Herzog & de Meuron.

Methods

The authors propose PCG-RAG, which uses Progressive Compositional Grounding to solve this. The framework consists of three main modules:

  1. Progressive Evidence Acquisition: Instead of accepting a candidate entity immediately, the system treats it as a "provisional" hypothesis. It identifies which constraints are still "unresolved" (e.g., "did this station open as a museum in 2000?") and specifically retrieves evidence to fill those gaps.
  2. Compositional Reasoning: The system decomposes the main question into a Directed Acyclic Graph (DAG) of subquestions. It uses provisional entity names to make subsequent searches highly specific (e.g., "Architect of [Candidate Name]"), but it always evaluates the results against the original identifying conditions to ensure the entity is still the right one.
  3. Evidence Grounding: During final answer generation, the system doesn't just provide the answer; it organizes the "source chain." It ensures that the final context provided to the LLM includes both the evidence identifying the entity and the evidence for the requested relation, preserving the logical link between them.
Figure 2: Overview of PCG-RAG, illustrated with a question about a project leader’s advisor and the advisor’s current affiliation. Progressive evidence acquisition updates entity hypotheses as missing evidence is retrieved. Compositional reasoning instanti- ates dependent queries along the subquestion DAG. Identifying and relational passages jointly support final answer generation.

Datasets

The authors evaluated PCG-RAG on three challenging benchmarks:

  • BrowseComp-Plus: A benchmark for deep search agents requiring multi-clue entity identification across a large web corpus.
  • WebWalkerQA: A dataset focused on navigating and searching through real-world web page structures.
  • MuSiQue: A multi-hop reasoning dataset used specifically to measure "support coverage"—how well the system finds the entire chain of necessary evidence.

Key Insights and Ideas

  • Provisionality is Key: Distinguishing between a "provisional search subject" and an "accepted answer" allows the system to use uncertain information to guide search without letting it corrupt the final reasoning.
  • The Identity-to-Relation Chain: A retrieved fact is only "grounded" if the system can prove it belongs to the entity described in the original question. PCG-RAG explicitly maintains this connection across different web pages.
  • Efficiency through Focus: By focusing acquisition on unresolved conditions, PCG-RAG avoids the massive token costs of "brute-force" search agents while achieving higher accuracy.

Key Results

PCG-RAG demonstrated significant improvements over state-of-the-art baselines like WebThinker, LQT, and DeepHone:

  • Accuracy: Improved accuracy by 5.43 percentage points on BrowseComp-Plus and 5.15 percentage points on WebWalkerQA compared to the strongest baselines.
  • Efficiency: On BrowseComp-Plus, it reduced token consumption by 82.98% and latency by 62.31% relative to WebThinker.
  • Evidence Discovery: On the MuSiQue dataset, it achieved a significantly higher "All@5" score (64.30%), meaning it was much more successful at retrieving the complete set of supporting passages required for a multi-hop answer.
Figure 3: New evidence connects identifying clues to correct downstream answers on BrowseComp-Plus. The displayed early entity records remain unchanged.

Limitations and Future Work

  • Reasoning Depth: While PCG-RAG leads at all depths, its performance (like all current RAG systems) degrades as the number of reasoning "hops" increases. At 4 hops, the "All@5" evidence recovery drops to 16.87%.
  • Static Planning: The current subquestion DAG is constructed at the start. Future work could involve more dynamic re-planning where the system can restructure its reasoning graph based on the evidence it discovers mid-search.
  • Computational Overhead: While more efficient than some "agentic" searchers, the multi-step nature still involves multiple LLM calls, which may be a constraint for real-time applications.

Summary updated

Abstract

Complex Web question answering requires retrieval-augmented generation (RAG) systems to identify target entities from multi-clue descriptions and connect evidence scattered across long Web pages. Individual pages often support only some of the required semantic constraints, while clues discovered on one page guide subsequent searches on others. However, premature reliance on partially supported entity hypotheses can divert retrieval from the original information need. RAG systems must therefore use partial evidence to guide further searches while seeking the evidence needed to establish the intended entities. We propose PCG-RAG, a novel RAG framework based on Progressive Compositional Grounding (PCG). Progressive Evidence Acquisition directs subsequent searches toward missing evidence for candidate attributes and relations. Compositional Reasoning separates candidate-guided retrieval from evidence-based answering along a directed acyclic graph (DAG) of subquestions. Provisional names make dependent searches specific, while answering tasks use accepted results and unresolved identifying conditions, assessed against the original question. New passages connect the identifying clues to the requested relations, supporting evidence-based downstream answers. Evidence Grounding combines identifying and relational source passages with supported intermediate answers, connecting evidence across Web pages to the original conditions during final generation. On BrowseComp-Plus and WebWalkerQA, PCG-RAG improves accuracy over the leading compared baselines by 5.43 and 5.15 percentage points, respectively.
BibTeX
@misc{zheng2026pcg,
  author = {Linyao Zheng and Xuhang Shi and Zhifang Mao and Chongqi Qu and Xiuquan Hou},
  title = {{PCG-RAG: Progressive Compositional Grounding for Retrieval-Augmented Generation in Complex Web Question Answering}},
  year = {2026},
  eprint = {2611.11113},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2611.11113},
}
Figure