Multi-Turn Agentic Scientific Literature Search via Workflow Induction
For researchers conducting complex literature searches, PaperPilot provides a controllable and inspectable multi-turn search agent that significantly outperforms baseline tool-use agents.
PaperPilot frames scientific literature search as workflow induction, constructing executable DAGs of search operators from an anchor paper and user query, then refining both query and workflow via user feedback. It achieves Hit@5 of 77.0 (vs. 58.0 baseline), MRR of 59.4 (vs. 47.5), and nDCG@10 of 32.5 (vs. 26.8), with 0% workflow errors.
Scientific literature search often requires more than retrieving papers from a single query: users' intents are underspecified, preference-dependent, and evolve through interaction. Existing search agents typically rely on fixed pipelines or implicit language-only reasoning, making their search strategies difficult to control, inspect, and refine. We introduce PaperPilot, a multi-turn literature search agent that frames scientific search as workflow induction. Given an anchor paper and a user query, PaperPilot constructs an executable DAG of paper-search operators, including keyword search, citation expansion, filtering, scoring, reranking, and evidence extraction. User feedback is then used to refine both the query and the workflow itself. We train PaperPilot with supervised workflow imitation and preference optimization over controlled workflow corruptions. Experiments show that PaperPilot-9B improves over the base Qwen3.5-9B toolset agent under multi-turn interaction, increasing Hit@5 from 58.0 to 77.0, MRR from 47.5 to 59.4, and nDCG@10 from 26.8 to 32.5, while reducing workflow execution errors from 9.5% to 0%. These results show that explicit, editable search workflows provide an effective and controllable interface for aligning literature search agents with complex scientific intent.