{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,8,2]],"date-time":"2025-08-02T16:24:23Z","timestamp":1754151863081,"version":"3.41.2"},"reference-count":0,"publisher":"Association for the Advancement of Artificial Intelligence (AAAI)","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["SOCS"],"abstract":"<jats:p>Reinforcement learning is becoming ever more prominent in solving combinatorial search problems, in particular ones where states are images. Prior work has devised action-policy testing methodology, that identifies so-called bug states where policy performance is sub-optimal. Here we show how to leverage this methodology during the RL process, using action-policy testing to find bugs and injecting those as alternate start states for the training runs. Running experiments across six 2D games, we find that our testing-guided training often achieves similar expected reward while reducing the number of bugs.<\/jats:p>","DOI":"10.1609\/socs.v18i1.35990","type":"journal-article","created":{"date-parts":[[2025,7,21]],"date-time":"2025-07-21T06:05:14Z","timestamp":1753077914000},"page":"181-185","source":"Crossref","is-referenced-by-count":0,"title":["Using Action-Policy Testing in RL to Reduce the Number of Bugs"],"prefix":"10.1609","volume":"18","author":[{"given":"Hasan Ferit","family":"Eniser","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Songtuan","family":"Lin","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Nicola","family":"M\u00fcller","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Anastasia","family":"Isychev","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Valentin","family":"W\u00fcstholz","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Isabel","family":"Valera","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"J\u00f6rg","family":"Hoffmann","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Maria","family":"Christakis","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"9382","published-online":{"date-parts":[[2025,7,19]]},"container-title":["Proceedings of the International Symposium on Combinatorial Search"],"original-title":[],"link":[{"URL":"https:\/\/ojs.aaai.org\/index.php\/SOCS\/article\/download\/35990\/38145","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/ojs.aaai.org\/index.php\/SOCS\/article\/download\/35990\/38145","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,7,21]],"date-time":"2025-07-21T06:05:14Z","timestamp":1753077914000},"score":1,"resource":{"primary":{"URL":"https:\/\/ojs.aaai.org\/index.php\/SOCS\/article\/view\/35990"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7,19]]},"references-count":0,"URL":"https:\/\/doi.org\/10.1609\/socs.v18i1.35990","relation":{},"ISSN":["2832-9163","2832-9171"],"issn-type":[{"type":"electronic","value":"2832-9163"},{"type":"print","value":"2832-9171"}],"subject":[],"published":{"date-parts":[[2025,7,19]]}}}