Collaborative Research: Conference: Large Language Models for Biological Discoveries (LLMs4Bio)

大型语言模型（llm）的计算研究正在迅速发展，并扩展到自然语言处理（NLP）之外。特别是，人们对如何利用和改进这些模型以实现跨科学学科的科学探究非常感兴趣。该奖项将支持组织一个专注于利用llm进行生物学发现的研讨会：LLMs4Bio将伴随2024年AAAI会议。AAAI系列会议已成为世界上首屈一指的人工智能研究会议。本次研讨会解决了这些挑战，并汇集了来自计算机科学，信息科学，分子，细胞和系统生物学的研究人员，以解决推进生物学发现的独特挑战。该研讨会活动将通过向美国学生提供旅行资助，提高学生对将于2月22日至28日在加拿大温哥华举行的2024年AAAI人工智能会议（AAAI）的参与度。AAAI系列会议已成为世界上首屈一指的人工智能研究会议。其成果包括制定新的问题空间，将更多的研究人员纳入已确定的交叉社区，以及促进可访问和包容性法学硕士的进一步创新，以推动下一个科学突破。美国研究人员在会议上的强大代表也有助于保持美国在这一重要领域的竞争力。本次研讨会活动将提供一个展示人工智能技术进步和在社会领域应用的国际论坛。会议涵盖了人工智能的各个方面，包括理论、算法、软件和系统以及应用。该奖项将支持：i)“生物学法学硕士”小组讨论，ii)论文展示和问答互动，iii) NLP和生物领域专家之间的网络，以及iv)我们社区未来机会的展望。这些活动将支持学生和年轻研究人员准备在科学研究领域推进他们的职业生涯，并扩大代表性不足的群体在计算领域，特别是在人工智能研究领域的参与。法学硕士目前的研究速度使研究人员难以深入理解科学问题、社区可接受的标准、数据集、指标和基准任务，这些任务真正体现了我们在问题上取得进展的能力，以及数十年艰苦研究中收集的宝贵知识。目前的趋势是，封闭或描述不清的行业模型超出了学术研究人员通常可以获得的资源，通常通过非同行评审平台传播，这也不利于研究的交叉受精。本次研讨会解决了这些挑战，并汇集了来自计算机科学、信息科学、分子、细胞和系统生物学的研究人员，以解决推进生物学发现的独特挑战，如标准化数据集、社区接受的基准、实验噪声和不确定性量化、解释和先前生物学知识的注入。与研讨会活动相伴随的Github项目https://github.com/LLMs4Science-Community提供了一个长期平台，用于在本次首次提供的产品和其他计划的年度产品中共享研讨会研究文章、数据集、基准、指标和其他相关知识。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Computational research on large language models (LLMs) is advancing rapidly and expanding beyond Natural Language Processing (NLP). In particular, there is great interest in how these models can be leveraged and advanced to enable scientific inquiry across scientific disciplines. This award will support the organization of a workshop focused on the utilization of LLMs for biological discoveries: LLMs4Bio accompanies the 2024 AAAI conference. The AAAI conference series has established itself as the world’s premier research conference in AI. This workshop addresses these challenges and brings together researchers from computer science, information science, and molecular, cellular, and systems biology to address unique challenges in advancing biological discoveries. This workshop activity will increase student participation in the 2024 AAAI Conference on Artificial Intelligence (AAAI) that will take place from Feb 22-28 in Vancouver, Canada, by providing travel grants to U.S.-based students. The AAAI conference series has established itself as the world’s premier research conference in AI. Outcomes include the formulation of new problem spaces, the inclusion of more researchers in the identified intersectional communities, and the catalysis of further innovation on accessible and inclusive LLMs to power the next scientific breakthroughs. A strong representation of US researchers at the conference also helps maintain US competitiveness in this important area.This workshop activity will provide an international forum for presentation of AI technological advancement and applications in societal areas. The conference covers all aspects of AI, including theories, algorithms, software, and systems, and applications. The award will support: i) “LLMs for biology” panel discussions, ii) paper presentation and Q&A interactions, iii) networking across NLP and biological domain experts, and iv) outlook of future opportunities of our communities. These activities will support students and young researchers as they prepare to advance their careers in scientific research, as well as broaden the participation of under-represented groups in computing and, in particular, in AI research. The current pace of research in LLMs makes it challenging for researchers to deeply understand scientific problems, community-acceptable standards, datasets, metrics, and benchmark tasks that truly capture our ability to advance on a problem, and precious knowledge gathered over decades of hard-fought research. The current trend of closed or poorly-described industry models that remain beyond resources typically available to academic researchers, often disseminated through non-peer reviewed platforms is also not conducive to cross-fertilization of research. This workshop addresses these challenges and brings together researchers from computer science, information science, and molecular, cellular, and systems biology to address unique challenges in advancing biological discoveries, such as standardized datasets, community-accepted benchmarks, experimental noise and uncertainty quantification, interpretation, and injection of prior biological knowledge. A Github project, https://github.com/LLMs4Science-Community, that accompanies the workshop activity provides a long-term platform for sharing workshop research articles, datasets, benchmarks, metrics, and other resulting knowledge for the workshop in this debut offering and other planned annual offerings.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.