{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/atomic-an-atlas-of-machine-commonsense-for-if","title":"ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning","arxiv_id":"1811.00146","date":"2018-10-31","proceeding":null,"authors":["Maarten Sap","Ronan LeBras","Emily Allaway","Chandra Bhagavatula","Nicholas Lourie","Hannah Rashkin","Brendan Roof","Noah A. Smith","Yejin Choi"],"abstract":"We present ATOMIC, an atlas of everyday commonsense reasoning, organized\nthrough 877k textual descriptions of inferential knowledge. Compared to\nexisting resources that center around taxonomic knowledge, ATOMIC focuses on\ninferential knowledge organized as typed if-then relations with variables\n(e.g., \"if X pays Y a compliment, then Y will likely return the compliment\").\nWe propose nine if-then relation types to distinguish causes vs. effects,\nagents vs. themes, voluntary vs. involuntary events, and actions vs. mental\nstates. By generatively training on the rich inferential knowledge described in\nATOMIC, we show that neural models can acquire simple commonsense capabilities\nand reason about previously unseen events. Experimental results demonstrate\nthat multitask models that incorporate the hierarchical structure of if-then\nrelation types lead to more accurate inference compared to models trained in\nisolation, as measured by both automatic and human evaluation.","url_abs":"http://arxiv.org/abs/1811.00146v3","url_pdf":"http://arxiv.org/pdf/1811.00146v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"atomic-an-atlas-of-machine-commonsense-for-if","repo_url":"https://github.com/AMGrobelnik/MultiCOMET","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"atomic-an-atlas-of-machine-commonsense-for-if","repo_url":"https://github.com/simon-benigeri/narrative-generation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":null,"task_name":"Relation"}],"methods":[],"datasets_introduced":[{"slug":"atomic","name":"ATOMIC","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.00146","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}