{"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/uncertainty-aware-attention-for-reliable","title":"Uncertainty-Aware Attention for Reliable Interpretation and Prediction","arxiv_id":"1805.09653","date":"2018-05-24","proceeding":"NeurIPS 2018 12","authors":["Jay Heo","Hae Beom Lee","Saehoon Kim","Juho Lee","Kwang Joon Kim","Eunho Yang","Sung Ju Hwang"],"abstract":"Attention mechanism is effective in both focusing the deep learning models on\nrelevant features and interpreting them. However, attentions may be unreliable\nsince the networks that generate them are often trained in a weakly-supervised\nmanner. To overcome this limitation, we introduce the notion of input-dependent\nuncertainty to the attention mechanism, such that it generates attention for\neach feature with varying degrees of noise based on the given input, to learn\nlarger variance on instances it is uncertain about. We learn this\nUncertainty-aware Attention (UA) mechanism using variational inference, and\nvalidate it on various risk prediction tasks from electronic health records on\nwhich our model significantly outperforms existing attention models. The\nanalysis of the learned attentions shows that our model generates attentions\nthat comply with clinicians' interpretation, and provide richer interpretation\nvia learned variance. Further evaluation of both the accuracy of the\nuncertainty calibration and the prediction performance with \"I don't know\"\ndecision show that UA yields networks with high reliability as well.","url_abs":"http://arxiv.org/abs/1805.09653v1","url_pdf":"http://arxiv.org/pdf/1805.09653v1.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":"uncertainty-aware-attention-for-reliable","repo_url":"https://github.com/OpenXAIProject/UncertintyAttention_DropMax","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"uncertainty-aware-attention-for-reliable","repo_url":"https://github.com/jayheo/UA","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"prediction","task_name":"Prediction"},{"task_slug":"variational-inference","task_name":"Variational Inference"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1805.09653","atlas_url":"https://app.syntology.ai/?focus=1805.09653","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}