Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Rethinking Deep Research from the Perspective of Web Content Distribution Matching

About

Despite the integration of search tools, Deep Search Agents often suffer from a misalignment between reasoning-driven queries and the underlying web indexing structures. Existing frameworks treat the search engine as a static utility, leading to queries that are either too coarse or too granular to retrieve precise evidence. We propose WeDas, a Web Content Distribution Aware framework that incorporates search-space structural characteristics into the agent's observation space. Central to our method is the Query-Result Alignment Score, a metric quantifying the compatibility between agent intent and retrieval outcomes. To overcome the intractability of indexing the dynamic web, we introduce a few-shot probing mechanism that iteratively estimates this score via limited query accesses, allowing the agent to dynamically recalibrate sub-goals based on the local content landscape. As a plug-and-play module, WeDas consistently improves sub-goal completion and accuracy across four benchmarks, effectively bridging the gap between high-level reasoning and low-level retrieval.

Zixuan Yu, Zhenheng Tang, Tongliang Liu, Chengqi Zhang, Xiaowen Chu, Bo Han• 2026

Related benchmarks

TaskDatasetResultRank
Data Science Agent tasksxBench-DS
Pass@172
31
General AI Assistant ReasoningBrowseComp-zh (BC-zh)
Pass@1 Accuracy41
19
General AI Assistant ReasoningBrowseComp (BC)
Pass@1 Accuracy26
17
General AI Assistant ReasoningGAIA
Pass@1 Accuracy66.99
17
Web Browsing ResearchBrowseComp (BC)
Pass@335
6
Web Browsing ResearchBrowseComp-zh (BC-zh)
Pass@358
6
Web Browsing ResearchGAIA
Pass@375.73
6
Web Browsing ResearchXBench-DS (XBD)
Pass@386
6
Showing 8 of 8 rows

Other info

Follow for update