Open-source AI tool beats giant LLMs in literature reviews — and gets citations right开源AI工具在文献综述方面击败了大型LLMs，并正确引用

Researchers have published the recipe for an artificial-intelligence model that reviews the scientific literature better than some major large language models (LLMs) are able to, and gets the citations correct as often as human experts do.研究人员已经发布了一种人工智能模型的配方，它比一些主要的大型语言模型（llm）更好地审查科学文献，并像人类专家一样经常得到正确的引用。

OpenScholar — which combines a language model with a database of 45 million open-access articles — links the information it sources directly back to the literature, to stop the system from making up or ‘hallucinating’ citations.OpenScholar将一个语言模型和一个包含4500万篇开放获取文章的数据库结合起来，将它所获取的信息直接链接到文献上，以防止系统编造或夸大引用。

Several commercial AI-based literature-review tools already exist that use similar techniques, but few have been released as open source, says Akari Asai, an AI researcher at Carnegie Mellon University in Pittsburgh, Pennsylvania, and a co-author of the work, published in Nature on 4 February¹. Being open source means that researchers can not only try OpenScholar for free in an online demonstration, but also deploy it on their own machine and use the method in the paper to boost the literature-review skills of any LLM, says Asai.宾夕法尼亚州匹兹堡卡耐基梅隆大学的人工智能研究员Akari Asai说，一些基于人工智能的商业文献综述工具已经存在，它们使用类似的技术，但很少以开源方式发布，2月4日发表的这项工作的合著者<；b0>；开源意味着研究人员不仅可以在在线演示中免费尝试OpenScholar，还可以在自己的机器上部署它，并使用论文中的方法来提高任何LLM的文献综述技能，Asai说。

In the 14 months since OpenScholar was first published in the arXiv repository², AI firms such as OpenAI have used similar methods to tack ‘deep research’ tools onto their commercial LLMs, which has greatly improved their accuracy. But as a small and efficient system, running OpenScholar costs a fraction of the price of using OpenAI’s GPT-5 with deep research, co-author Hannaneh Hajishirzi, a computer scientist at the University of Washington in Seattle, tells the Nature podcast.在OpenScholar首次发表在arXiv仓库的14个月以来，OpenAI等人工智能公司使用类似的方法将深度研究工具添加到他们的商业llm中，这大大提高了它们的准确性。但是，作为一个小型而高效的系统，运行OpenScholar的成本只是使用openai的GPT-5进行深入研究的成本的一小部分，西雅图华盛顿大学的计算机科学家、合著者Hannaneh Hajishirzi在播客上说。

However, the authors acknowledge that OpenScholar has limitations. For example, it doesn’t always retrieve the most representative or relevant papers for a query, and it is limited by the scope of its database.然而，作者承认OpenScholar有局限性。例如，它并不总是为查询检索到最具代表性或最相关的论文，并且受到其数据库范围的限制。

But if researchers are able to access the tool for free, “it can become one of the most popular apps for scientific searches,” says Mushtaq Bilal, a researcher at Silvi, a Copenhagen-based firm that has its own AI-based literature-review tool.但如果研究人员能够免费使用该工具，它将成为最受欢迎的科学搜索应用程序之一，Mushtaq Bilal说，他是Silvi公司的研究员，这家总部位于哥本哈根的公司拥有自己的基于人工智能的文献综述工具。

Outperforming humans?

超越人类?

LLMs can write fluently, but they often struggle with citations. This is because they learn by building links between words in their training data, which include sources outside science, and then generate text on the basis of probable associations that are not always correct or up to date. This is a feature of LLMs, not a bug, and it is proving to be a problem when people use LLMs in research. For example, at least 51 papers accepted to the high-profile machine learning NeurIPS conference in December 2025, contained non-existent or inaccurate citations, according to an analysis using the GPTZero tool.llm可以流利地写作，但他们经常在引用方面遇到困难。这是因为他们通过在训练数据（包括科学以外的来源）中建立单词之间的联系来学习，然后根据并不总是正确或最新的可能联系生成文本。这是llm的一个特性，而不是bug，并且当人们在研究中使用llm时，它被证明是一个问题。例如，根据使用GPTZero工具的分析，在2025年12月备受瞩目的机器学习NeurIPS会议上，至少有51篇论文被接受，其中包含不存在或不准确的引用。

OpenScholar is a way to force an LLM to answer queries using a specific data store, says Hajishirzi. When a user asks a question, a retrieval system finds related scientific articles in the repository, then ranks them by relevance and generates a response that is based only on the most useful. The LLM — trained on examples of questions and answers — then refines the answer to improve it. “We designed an efficient pipeline where the model generates an answer once, but then keeps improving if needed,” says Asai.Hajishirzi说，OpenScholar是一种强制LLM使用特定数据存储来回答查询的方法。当用户提出一个问题时，检索系统会在存储库中找到相关的科学文章，然后根据相关性对它们进行排序，并只根据最有用的来生成响应。LLM通过问题和答案的例子进行训练，然后提炼答案以改进它。Asai说：“我们设计了一个高效的流水线，模型只产生一次答案，然后在需要的时候继续改进。”

Although the data store used in the paper contains scientific articles up to October 2024, Asai says that the demo version of OpenScholar can tap into the academic search engine Semantic Scholar, from the Allen Institute for Artificial Intelligence in Seattle, to access current papers. And because responses are built around real, specific papers, the system very rarely fabricates citations. It can, however, still cite a paper that doesn’t support a claim very well, in the same way that humans can, says Asai.虽然论文中使用的数据存储包含到2024年10月的科学文章，Asai表示，OpenScholar的演示版本可以访问来自西雅图艾伦人工智能研究所的学术搜索引擎Semantic Scholar，以访问当前的论文。而且，由于反馈是围绕真实的、具体的论文建立的，该系统很少捏造引文。但是，它仍然可以引用一篇不能很好地支持某个主张的论文，就像人类可以那样，Asai说。

To test OpenScholar, the team compared its ability to answer realistic queries with that of other AI tools and with answers written by experienced, PhD-level humans, in computer science, physics, neuroscience and biomedicine. The team found that experts preferred the models’ responses over human-written ones in most cases. OpenScholar also outperformed GPT-4o, as well as rival tools such as PaperQA2, from FutureHouse in San Francisco, California, on machine-judged evaluations of citation and factual accuracy.为了测试OpenScholar，该团队将其回答现实问题的能力与其他AI工具的能力以及由计算机科学、物理、神经科学和生物医学领域有经验的博士水平的人类撰写的答案进行了比较。研究小组发现，在大多数情况下，专家更喜欢模型反应，而不是人工编写的反应。OpenScholar在引用和事实准确性的机器判断评估方面也优于GPT-4o以及来自加利福尼亚州旧金山FutureHouse的PaperQA2等竞争工具。

A limitation of the tool is that it cannot ensure that it accesses only scientifically rigorous articles, says Bilal. It also cannot access paywalled papers, which means it is much less useful in disciplines, such as engineering or social sciences, in which open-access preprints are not the norm, he says. “Lack of access to copyrighted, licensed, or paywalled research is one of the biggest bottlenecks in improving these AI tools,” he says.Bilal说，这个工具的一个限制是它不能确保它只获取科学严谨的文章。他说，它也不能访问付费的论文，这意味着它在工程学或社会科学等学科中用处不大，这些学科的开放获取预印本不是常态。他说，缺乏获得版权、许可或付费研究的途径是改进这些人工智能工具的最大瓶颈之一。

In the future, the team would like to develop a more flexible system that allows users to tap into papers they hold subscriptions for and that they have downloaded locally, says Asai.Asai说，未来，该团队希望开发一个更灵活的系统，允许用户使用他们订阅的报纸，并在本地下载。

https://www.nature.com/articles/d41586-026-00347-9