{"id":408,"date":"2025-03-25T22:15:53","date_gmt":"2025-03-25T22:15:53","guid":{"rendered":"https:\/\/techtrendfeed.com\/?p=408"},"modified":"2025-03-25T22:15:53","modified_gmt":"2025-03-25T22:15:53","slug":"exploring-empty-areas-human-in-the-loop-knowledge-augmentation","status":"publish","type":"post","link":"https:\/\/techtrendfeed.com\/?p=408","title":{"rendered":"Exploring Empty Areas: Human-in-the-Loop Knowledge Augmentation"},"content":{"rendered":"<p> <br \/>\n<\/p>\n<div>\n<p>Knowledge augmentation is essential to make machine studying fashions extra strong and protected. Nevertheless, augmenting information may be difficult because it requires producing numerous information factors to carefully consider mannequin habits on edge circumstances and mitigate potential harms. Creating high-quality augmentations that cowl these &#8220;unknown unknowns&#8221; is a time- and creativity-intensive activity. On this work, we introduce Amplio, an interactive instrument to assist practitioners navigate &#8220;unknown unknowns&#8221; in unstructured textual content datasets and enhance information variety by systematically figuring out empty information areas to discover. Amplio contains three human-in-the-loop information augmentation methods: Increase with Ideas, Increase by Interpolation, and Increase with Massive Language Mannequin. In a person examine with 18 skilled purple teamers, we exhibit the utility of our augmentation strategies in serving to generate high-quality, numerous, and related mannequin security prompts. We discover that Amplio enabled purple teamers to enhance information rapidly and creatively, highlighting the transformative potential of interactive augmentation workflows.<\/p>\n<figure id=\"figure1\" class=\"\" aria-label=\"Figure 1\">\n<div class=\"bg-gray-light text-base rounded\"><a rel=\"nofollow\" target=\"_blank\" href=\"https:\/\/mlr.cdn-apple.com\/media\/teaser_8f97a249b4.png\" tabindex=\"-1\" target=\"_blank\" class=\"mt-0\"><img decoding=\"async\" src=\"https:\/\/mlr.cdn-apple.com\/media\/teaser_8f97a249b4.png\" loading=\"lazy\" class=\"bg-gray-light\"\/><\/a><\/div><figcaption class=\"muted\" id=\"figure-figure1-caption\" aria-hidden=\"true\">Determine 1: Given a dataset of unstructured textual content, it may be difficult to find out how and the place to enhance the info most successfully. We suggest a visualization-based strategy to assist customers discover related empty information areas to discover to enhance dataset variety. To fill in these empty areas, metaphorically represented by gaps in an embedding plot, we design an interactive instrument with three human-in-the-loop augmentation strategies: Increase with Ideas, Increase by Interpolation, and Increase with Massive Language Mannequin. Right here, every dot represents an embedded sentence from the enter dataset of CHI 2024 paper titles.<\/figcaption><\/figure>\n<\/div>\n\n","protected":false},"excerpt":{"rendered":"<p>Knowledge augmentation is essential to make machine studying fashions extra strong and protected. Nevertheless, augmenting information may be difficult because it requires producing numerous information factors to carefully consider mannequin habits on edge circumstances and mitigate potential harms. Creating high-quality augmentations that cowl these &#8220;unknown unknowns&#8221; is a time- and creativity-intensive activity. On this work, [&hellip;]<\/p>\n","protected":false},"author":2,"featured_media":410,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[55],"tags":[158,157,154,79,156,155],"class_list":["post-408","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-machine-learning","tag-augmentation","tag-data","tag-empty","tag-exploring","tag-humanintheloop","tag-spaces"],"_links":{"self":[{"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/posts\/408","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=408"}],"version-history":[{"count":1,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/posts\/408\/revisions"}],"predecessor-version":[{"id":409,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/posts\/408\/revisions\/409"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=\/wp\/v2\/media\/410"}],"wp:attachment":[{"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=408"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=408"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/techtrendfeed.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=408"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}<!-- This website is optimized by Airlift. Learn more: https://airlift.net. Template:. Learn more: https://airlift.net. Template: 69d9690a190636c2e0989534. Config Timestamp: 2026-04-10 21:18:02 UTC, Cached Timestamp: 2026-06-27 20:13:23 UTC -->