跳到正文
arXiv cs.AI· Frank Nie, Ethan B. Liu, Yuan Zhu, Jindong Han·· 2 小时前AI 评分34

协同证据检索与利用的自改进多模态科学文档推理框架

Synergizing Evidence Retrieval and Utilization for Self-Improving Multimodal Scientific Document Reasoning

AI 导读

一项 arXiv 研究提出面向多模态科学文档推理的 Retriever–Perceiver–Reasoner 智能体框架:Retriever 在上下文预算内顺序选择证据,Perceiver 构建问题条件的潜在本体,Reasoner 结合原始来源生成有依据的答案。

正文 · 原文

View PDF HTML (experimental)

Abstract:Scientific document reasoning requires an agent to acquire and integrate evidence distributed across text, figures, and tables. Retrieval can identify relevant sources while missing complementary support, and answer-level feedback alone does not reveal whether an error stems from evidence acquisition or evidence use. We present an agentic Retriever--Perceiver--Reasoner framework for multimodal scientific document reasoning. The Retriever sequentially selects source-aligned evidence under a context budget. Given the selected evidence and question, the Perceiver constructs a question-conditioned latent ontology, an implicit organization of multimodal evidence in a shared continuous space. The Reasoner combines this latent guidance with the original sources to produce grounded answers. We train the three components through diagnostic-guided co-evolution: alternating updates optimize them with a shared answering reward, while Retrieval Potential Feedback (RPF) and Latent Efficacy Feedback (LEF) identify questions that need better retrieval or latent guidance and direct targeted resampling. Across four benchmarks and seven evaluation settings, our framework obtains average scores of 69.9 and 75.0 with 7B and 8B Reasoner backbones, exceeding the strongest baseline by 4.0 and 9.1 points, respectively.
Subjects: Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.22837 [cs.AI]
  (or arXiv:2611.22837v1 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2611.22837

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Nie Frank [view email]
[v1] Tue, 17 Nov 2026 14:19:18 UTC (960 KB)

来源:arXiv cs.AI · arxiv.org