Search papers, labs, and topics across Lattice.
The paper introduces RECIPER, a dual-view retrieval pipeline for materials science QA that combines paragraph-level dense retrieval with retrieval over LLM-extracted procedural summaries. By indexing both paragraph context and concise procedural summaries, RECIPER significantly improves early-rank retrieval performance compared to paragraph-only methods. Experiments using BGE-large-en-v1.5 demonstrate substantial gains in Recall@1, nDCG@10, and MRR, leading to improved downstream question answering.
LLM-distilled procedural summaries, when combined with traditional dense retrieval, unlock significant gains in materials science question answering, outperforming paragraph-only retrieval.
Retrieving procedure-oriented evidence from materials science papers is difficult because key synthesis details are often scattered across long, context-heavy documents and are not well captured by paragraph-only dense retrieval. We present RECIPER, a dual-view retrieval pipeline that indexes both paragraph-level context and compact large language model-extracted procedural summaries, then combines the two candidate streams with lightweight lexical reranking. Across four dense retrieval backbones, RECIPER consistently improves early-rank retrieval over paragraph-only dense retrieval, achieving average gains of +3.73 in Recall@1, +2.85 in nDCG@10, and +3.13 in MRR. With BGE-large-en-v1.5, it reaches 86.82%, 97.07%, and 97.85% on Recall@1, Recall@5, and Recall@10, respectively. We further observe improved downstream question answering under automatic metrics, suggesting that procedural summaries can serve as a useful complementary retrieval signal for procedure-oriented materials question answering. Code and data are available at https://github.com/ReaganWu/RECIPER.
