AI RESEARCH
WebAggregator: Enhancing Compositional Reasoning Capabilities of Deep Research Agent Foundation Models
arXiv CS.CL
•
ArXi:2510.14438v2 Announce Type: replace The hallmark of Deep Research agents lies in compositional reasoning, the capacity to aggregate distributed, heterogeneous information into coherent logical insights. However, current agentic systems are often retrieval-heavy but reasoning-light, where success is predominantly determined by simple entity-seeking rather than the multi-step aggregation of scattered evidence. To address this, we propose a data synthesis pipeline WebAggregator, designed to shift the agentic paradigm from retrieval-centric to compositional aggregation.