arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2409.11971cs.CLcond-mat.mtrl-sci
收藏

从大语言模型生成的嵌入表示中采样潜在材料属性信息

Sampling Latent Material-Property Information From LLM-Derived Embedding Representations

Trinity College(都柏林三一学院);The University of Utah(犹他大学)

Luke P. J. Gilligan, Matteo Cobelli, Hasan M. Sayeed, Taylor D. Sparks, Stefano Sanvito

AI总结:

本研究探究大语言模型生成的向量嵌入捕获材料潜在属性信息的能力,发现其虽能生成反映特定属性的表示,但需匹配最优上下文线索和比较基准,仍具材料科学表示生成潜力。

摘要:

由大语言模型(LLM)生成的向量嵌入在捕获文献中的潜在信息方面展现出前景。值得注意的是,这些向量可整合到材料嵌入中,有望用于数据驱动的材料属性预测。我们探究了LLM生成的向量在多大程度上能捕获所需信息,以及它们在无需额外训练的情况下提供材料属性见解的潜力。研究结果表明,尽管LLM可用于生成反映特定属性信息的表示,但提取嵌入需要识别最优上下文线索和合适的比较基准。尽管存在这一限制,LLM似乎仍有望在生成有意义的材料科学表示方面发挥作用。

英文摘要:

Vector embeddings derived from large language models (LLMs) show promise in capturing latent information from the literature. Interestingly, these can be integrated into material embeddings, potentially useful for data-driven predictions of materials properties. We investigate the extent to which LLM-derived vectors capture the desired information and their potential to provide insights into material properties without additional training. Our findings indicate that, although LLMs can be used to generate representations reflecting certain property information, extracting the embeddings requires identifying the optimal contextual clues and appropriate comparators. Despite this restriction, it appears that LLMs still have the potential to be useful in generating meaningful materials-science representations.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑