AI Paper Atlas

Research map / Multi-agent systems and games / LLM agents

FlowSearcher: Synthesizing Memory-Guided Agentic Workflows for Web Information Seeking

Keyi Xiang, Zeyu Feng, Zhuoyi Lin, Yueming Lyu, Shi Boyuan, Yew-Soon Ong, Ivor Tsang, Haiyan Yin

ICLR 2026 · Multi-agent systems and games › LLM agents · approach: Agentic benchmarks and orchestration

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes FlowSearcher, a web information seeking framework that synthesizes memory-guided agentic workflows for deep research agents instead of fixed ReAct-style tool chains.

Problem

Existing web search agents follow rigid ReAct-style tool chains with fixed workflow structures, limiting flexibility across diverse query types and tool-use strategies.

Approach

Builds on agentic workflow synthesis guided by memory, so the agent composes workflow structures adaptively per query rather than following a fixed ReAct chain. Full details unavailable beyond the abstract.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
expand block library from First-Hit only to First-Hit + General (ablation)FlowSearcher (GPT-4o-mini) · GAIAPass@1 avg: 27.2 -> 35.0Improved
remove all block restrictions (add in-depth browsing and parallel/ensemble blocks) (ablation)FlowSearcher (GPT-4o-mini) · GAIAPass@1 avg: 35.0 -> 55.3Improved
add full hierarchical memory (ablation)FlowSearcher (GPT-4o-mini) · GAIA 103 taskssuccessful tasks: 42 -> 57Improved
store only successful episodes in memory (ablation)FlowSearcher (GPT-4o-mini) · GAIA tasks 1-40successful tasks: 18 -> 20 (at 40 tasks); 42 -> 53 (at 103)Mixed
store only unsuccessful episodes in memory (ablation)FlowSearcher (GPT-4o-mini) · GAIA 103 taskssuccessful tasks: 42 -> 48Improved
add RL fine-tuning to WebThinkerWebThinker (QwQ-32B) · GAIAPass@1 avg: 44.7 -> 48.5Improved
add RL fine-tuning to WebThinkerWebThinker (QwQ-32B) · BrowseCompPass@1 avg: 2.3 -> 2.7No clear effect
replace ReAct agent with FlowSearcher workflow synthesisQwQ-32B · BrowseCompPass@1 avg: 3.8 -> 8.0Improved
add search agency (vanilla ReAct) to Qwen-2.5-32BQwen-2.5-32B · GAIAPass@1 avg: 13.0 -> 31.0Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Experimental setup

Baselines: ReAct-style tool chains

Builds on: ReAct

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy