Supernotes: Driving Consensus in Crowd-Sourced Fact-Checking
Soham De, Michiel A. Bakker, Jay Baxter, Martin Saveski
Abstract
X's Community Notes, a crowd-sourced fact-checking system, allows users to annotate potentially misleading posts. Notes rated as helpful by a diverse set of users are prominently displayed below the original post. While demonstrably effective at reducing misinformation's impact when notes are displayed, there is an opportunity for notes to appear on many more posts: for 91% of posts where at least one note is proposed, no notes ultimately achieve sufficient support from diverse users to be shown on the platform. This motivates the development of Supernotes: AI-generated notes that synthesize information from several existing community notes and are written to foster consensus among a diverse set of users. Our framework uses an LLM to generate many diverse Supernote candidates from existing proposed notes. These candidates are then evaluated by a novel scoring model, trained on millions of historical Community Notes ratings, selecting candidates that are most likely to be rated helpful by a diverse set of users. To test our framework, we ran a human subjects experiment in which we asked participants to compare the Supernotes generated by our framework to the best existing community notes for 100 sample posts. We found that participants rated the Supernotes as significantly more helpful, and when asked to choose between the two, preferred the Supernotes 75.2% of the time. Participants also rated the Supernotes more favorably than the best existing notes on quality, clarity, coverage, context, and argumentativeness. Finally, in a follow-up experiment, we asked participants to compare the Supernotes against LLMgenerated summaries and found that the participants rated the Supernotes significantly more helpful, demonstrating that both the LLM-based candidate generation and the consensus-driven scoring play crucial roles in creating notes that effectively build consensus among diverse users. X Account @x_account BREAKING: In a stunning announcement, President Biden has announced that if he's reelected, he will cancel medical debt for millions of Americans. Let's go. POTUS does not have authority to cancel debt, SCOTUS ruled on previous attempt to cancel student loans. Biden would not be able to keep this campaign promise. https://www.npr.org/2023/06/30/1182216970/supreme-court-student- loan-forgiveness-decision-biden He's not going to do it. 1) He is currently the sitting president, being re-elected changes nothing 2) He does not have the power to do this Commonsense.com Biden's administration plans to ban medical debt from credit reports https://abcnews.go.com/Politics/sweeping-change-biden-administration-ban- medical-debt-credit/story?id=110997906
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Your agent calls
Luneget_paper_fulltext
Free to start. No credit card required.
Terminal
Install the CLIlune papers fulltext c5906d58-d745-45f3-9420-315726fcc61aCited by top-tier papers7
- Beyond the Crowd: LLM-Augmented Community Notes for Governing Health MisinformationJiaying Wu, Zihang Fu, Haonan Wang, Fanxiao Li et al.ACL 2026 · 17 citations
- Beyond Community Notes: A Framework for Understanding and Building Crowdsourced Context Systems for Social MediaTravis Lloyd, Tung Nguyen, Karen Levy, Mor NaamanCHI 2026 · 2 citations
- Value Profiles for Encoding Human VariationTaylor Sorensen, Pushkar Mishra, Roma Patel, Michael Henry Tessler et al.EMNLP 2025 · 2 citations
- Question the Questions: Auditing Representation in Online Deliberative ProcessesSoham De, Lodewijk Gelauff, Ashish Goel, Smitha Milli et al.WWW 2026 · 1 citation
- Request a Note: How the Request Function Shapes X's Community Notes SystemYuwei Chuai, Shuning Zhang, Ziming Wang, Xin Yi et al.CHI 2026 · 1 citation
Builds on6
- Jury Learning: Integrating Dissenting Voices into Machine Learning ModelsMitchell L. Gordon, Michelle S. Lam, Joon Sung Park, Kayur Patel et al.CHI 2022 · 134 citations
- Investigating Differences in Crowdsourced News Credibility Assessment: Raters, Tasks, and Expert CriteriaMd Momen Bhuiyan, Amy X. Zhang, Connie Moon Sehat, Tanushree MitraCSCW 2020 · 70 citations
- Will the Crowd Game the Algorithm?: Using Layperson Judgments to Combat Misinformation on Social Media by Downranking Distrusted SourcesZiv Epstein, Gordon Pennycook, David G. RandCHI 2020 · 68 citations
- Reinforcement Learning-based Counter-Misinformation Response Generation: A Case Study of COVID-19 Vaccine MisinformationBing He, Mustaque Ahamad, Srijan KumarWWW 2023 · 62 citations
- Sniffer: Multimodal Large Language Model for Explainable Out-of-Context Misinformation DetectionPeng Qi, Zehong Yan, Wynne Hsu, Mong-Li LeeCVPR 2024 · 54 citations
Related papers
- Gaming Consensus: Coordinated Manipulation in Crowdsourced Fact-CheckingNikil Selvam, Jay Baxter, Sophie Hilgard, Brad Miller et al.ICML 2026
- Did the Roll-Out of Community Notes Reduce Engagement With Misinformation on X/Twitter?Yuwei Chuai, Haoye Tian, Nicolas Pröllochs, Gabriele LenziniCSCW 2024 · 62 citations
- Community Fact-Checks Trigger Moral Outrage in Replies to Misleading Posts on Social MediaYuwei Chuai, Anastasia Sergeeva, Gabriele Lenzini, Nicolas PröllochsCHI 2025 · 5 citations
- Synthia: Scalable Grounded Persona Generation from Social Media DataVahid Rahimzadeh, Erfan Moosavi Monazzah, Mohammad Taher Pilehvar, Yadollah YaghoobzadehACL 2026 · 1 citation
- PoliCon: Evaluating LLMs on Achieving Diverse Political Consensus ObjectivesZhaowei Zhang, Xiaobo Wang, Minghua Yi, Mengmeng Wang et al.ICLR 2026 · 4 citations
