Large language models (LLMs) often generate content that contains factual errors when responding to fact-seeking prompts on open-ended topics. To benchmark a model's long-form factuality in open domains, we first use GPT-4 to generate LongFact, a prompt set comprising thousands of questions spanning 38 topics. We then propose that LLM agents can be used as automated evaluators for long-form factuality through a method which we call Search-Augmented Factuality Evaluator (SAFE). SAFE utilizes an LLM to break down a long-form response into a set of individual facts and to evaluate the accuracy of each fact using a multi-step reasoning process comprising sending search queries to Google Search and determining whether a fact is supported by the search results. Furthermore, we propose extending F1 score as an aggregated metric for long-form factuality. To do so, we balance the percentage of supported facts in a response (precision) with the percentage of provided facts relative to a hyperparameter representing a user's preferred response length (recall).
Empirically, we demonstrate that LLM agents can outperform crowdsourced human annotators - on a set of ~16k individual facts, SAFE agrees with crowdsourced human annotators 72% of the time, and on a random subset of 100 disagreement cases, SAFE wins 76% of the time. At the same time, SAFE is more than 20 times cheaper than human annotators. We also benchmark thirteen language models on LongFact across four model families (Gemini, GPT, Claude, and PaLM-2), finding that larger language models generally achieve better long-form factuality. LongFact, SAFE, and all experimental code are available at https://github.com/google-deepmind/long-form-factuality.
Beyond Under-Alignment: Atomic Preference Enhanced Factuality Tuning for Large Language Models
Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所);School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院)
FELM: Benchmarking Factuality Evaluation of Large Language Models
National University of Singapore(新加坡国立大学);The Hong Kong University of Science and Technology(香港科技大学);Carnegie Mellon University(卡内基梅隆大学);City University of Hong Kong(香港城市大学);Shanghai Jiao Tong University(上海交通大学)
FacLens: Transferable Probe for Foreseeing Non-Factuality in Fact-Seeking Question Answering of Large Language Models
Zhipu AI(智谱AI);Zhongguancun Laboratory(中关村实验室);Renmin University of China(中国人民大学);Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州));Tsinghua University(清华大学)
Mitigating Geospatial Knowledge Hallucination in Large Language Models: Benchmarking and Dynamic Factuality Aligning
College of AI, Tsinghua University(清华大学人工智能学院);Department of Electronic Engineering, BNRist, Tsinghua University(清华大学电子工程系);School of Electronic and Information Engineering, Beijing Jiaotong University(北京交通大学电子信息学院);Department of Computer Science and Technology, Tsinghua University(清华大学计算机科学与技术系)
Survey on Factuality in Large Language Models: Knowledge, Retrieval and Domain-Specificity
Westlake University(西湖大学);Zhejiang University(浙江大学);Purdue University(普渡大学);Fudan University(复旦大学);Amazon AWS AI(亚马逊AWS人工智能);Yale University(耶鲁大学);Shanghai Jiao Tong University(上海交通大学);The Hong Kong University of Science and Technology(香港科技大学);Tsinghua University(清华大学);Microsoft Research(微软研究院)
Beyond Factuality: A Comprehensive Evaluation of Large Language Models as Knowledge Generators
The Chinese University of Hong Kong(香港中文大学);National University of Singapore(新加坡国立大学);Tencent AI Lab(腾讯人工智能实验室);The Hong Kong University of Science and Technology(香港科技大学)
Hallucination to Truth: A Review of Fact-Checking and Factuality Evaluation in Large Language Models
Department of Computer Science and Engineering, United International University, Dhaka 1212, Bangladesh(乌姆特国际大学计算机科学与工程系);Department of Computer Science and Engineering, Daffodil International University, Dhaka-1341, Bangladesh(达芙尼国际大学计算机科学与工程系);Faculty of Science and Technology, Charles Darwin University, Casuarina, NT 0909, Australia(查尔斯·达尔文大学科学与技术学院)
Integrative Decoding: Improve Factuality via Implicit Self-consistency
The Hong Kong Polytechnic University(香港理工大学);Tsinghua University(清华大学);Microsoft Research(微软研究院);Microsoft Azure AI(微软Azure人工智能);University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校)
Factuality Challenges in the Era of Large Language Models
University of Copenhagen(哥本哈根大学);Mohamed bin Zayed University of Artificial Intelligence(穆罕默德·本·扎耶德人工智能大学);Korea Advanced Institute of Science and Technology(韩国科学技术院);Indian Institute of Technology Delhi(德里印度理工学院);University of Maryland(马里兰大学);Full Fact;Stanford University(斯坦福大学);University of Southern California(南加州大学);University of Oxford(牛津大学);Meta AI;Carnegie Mellon University(卡内基梅隆大学);University of Illinois Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);Indiana University(印第安纳大学);Newtrales;University of Wisconsin(威斯康星大学);Pagella Politica/Facta
Is Factuality Enhancement a Free Lunch For LLMs? Better Factuality Can Lead to Worse Context-Faithfulness
Institute of Computing Technology, CAS(中国科学院计算技术研究所);University of Chinese Academy of Sciences(中国科学院大学);University of California, Merced(加州大学默塞德分校);University of Science and Technology of China(中国科学技术大学)
FaStfact: Faster, Stronger Long-Form Factuality Evaluations in LLMs
UCLA(美国大学洛杉矶分校);CUHK(香港中文大学);HKUST (GZ)(香港科技大学(广州));HKUST(香港科技大学);Tsinghua University(清华大学);ECNU(华东师范大学);NVIDIA(英伟达);UCL(伦敦大学学院);HKU(香港大学);University of Cambridge(剑桥大学)