{"id":"https://openalex.org/W7203787672","doi":"https://doi.org/10.48550/arxiv.2608.17524","title":"Evaluating RL Explainability Methods by How Much They Help Fix Bugs in Agents","display_name":"Evaluating RL Explainability Methods by How Much They Help Fix Bugs in Agents","publication_year":2026,"publication_date":"2026-08-18","ids":{"openalex":"https://openalex.org/W7203787672","doi":"https://doi.org/10.48550/arxiv.2608.17524"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.17524","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.17524","pdf_url":"https://arxiv.org/pdf/2608.17524","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.17524","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5083920910","display_name":"Ram Rachum","orcid":"https://orcid.org/0009-0004-9278-7134","observed_orcids":["https://orcid.org/0009-0004-9278-7134"]},"institutions":[{"id":"https://openalex.org/I121934306","display_name":"Tufts University","ror":"https://ror.org/05wvpxv85","country_code":"US","type":"education","lineage":["https://openalex.org/I121934306"]},{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Rachum, Ram","raw_affiliation_strings":["Tufts University","University of California , Berkeley"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tufts University","institution_ids":["https://openalex.org/I121934306"]},{"raw_affiliation_string":"University of California , Berkeley","institution_ids":["https://openalex.org/I95457486"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034998890","display_name":"Yotam Amitai","orcid":"https://orcid.org/0000-0002-0084-9739","observed_orcids":["https://orcid.org/0000-0002-0084-9739"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Amitai, Yotam","raw_affiliation_strings":["Independent Researcher"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Independent Researcher","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069295532","display_name":"B\u00e1lint Gyevn\u00e1r","orcid":"https://orcid.org/0000-0003-0630-4315","observed_orcids":["https://orcid.org/0000-0003-0630-4315"]},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gyevn\u00e1r, B\u00e1lint","raw_affiliation_strings":["Carnegie Mellon University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Carnegie Mellon University","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036081979","display_name":"Reuth Mirsky","orcid":"https://orcid.org/0000-0003-1392-9444","observed_orcids":["https://orcid.org/0000-0003-1392-9444"]},"institutions":[{"id":"https://openalex.org/I121934306","display_name":"Tufts University","ror":"https://ror.org/05wvpxv85","country_code":"US","type":"education","lineage":["https://openalex.org/I121934306"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mirsky, Reuth","raw_affiliation_strings":["Tufts University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tufts University","institution_ids":["https://openalex.org/I121934306"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5111222160","display_name":"Cameron Allen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I95457486","display_name":"University of California, Berkeley","ror":"https://ror.org/01an7q238","country_code":"US","type":"education","lineage":["https://openalex.org/I95457486"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Allen, Cameron","raw_affiliation_strings":["University of California , Berkeley"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California , Berkeley","institution_ids":["https://openalex.org/I95457486"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":1.0,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":9.999999747378752e-05,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10743","display_name":"Software Testing and Debugging Techniques","score":0.0,"subfield":{"id":"https://openalex.org/subfields/1712","display_name":"Software"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/explainable-reinforcement-learning","display_name":"explainable reinforcement learning","score":0.9980000257492065},{"id":"https://openalex.org/keywords/bug-fixing","display_name":"bug fixes","score":0.42500001192092896},{"id":"https://openalex.org/keywords/closed-loop-evaluation","display_name":"closed-loop evaluation","score":0.675000011920929},{"id":"https://openalex.org/keywords/self-assessment","display_name":"self-assessment","score":0.5080000162124634}],"concepts":[{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.6579999923706055},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6567999720573425},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6413000226020813},{"id":"https://openalex.org/C179518139","wikidata":"https://www.wikidata.org/wiki/Q5140297","display_name":"Coding (social sciences)","level":2,"score":0.47119998931884766},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4706999957561493},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3855000138282776},{"id":"https://openalex.org/C3017944768","wikidata":"https://www.wikidata.org/wiki/Q1450463","display_name":"Poison control","level":2,"score":0.2842999994754791},{"id":"https://openalex.org/C140479938","wikidata":"https://www.wikidata.org/wiki/Q5254619","display_name":"Iterated function","level":2,"score":0.27970001101493835},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.2612000107765198},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.22460000216960907}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.17524","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.17524","pdf_url":"https://arxiv.org/pdf/2608.17524","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.17524","doi":"https://doi.org/10.48550/arxiv.2608.17524","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.17524","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.17524","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.17524","pdf_url":"https://arxiv.org/pdf/2608.17524","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7203787672.pdf","grobid_xml":"https://content.openalex.org/works/W7203787672.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4407091688","https://openalex.org/W4401857392","https://openalex.org/W7203667140","https://openalex.org/W4389131286","https://openalex.org/W4415961037","https://openalex.org/W7155021021","https://openalex.org/W7171424630","https://openalex.org/W7080326555","https://openalex.org/W4415309496","https://openalex.org/W7105892163"],"abstract_inverted_index":{"This":[0,150],"preliminary":[1],"paper":[2],"outlines":[3],"a":[4,58,62,75,109,157],"planned":[5],"evaluation":[6],"benchmark":[7,59,88],"for":[8,112,146],"Explainable":[9],"Reinforcement":[10],"Learning":[11],"(XRL)":[12],"methods.":[13],"Current":[14],"evaluations":[15],"rely":[16],"on":[17,25,134],"functionally-grounded":[18],"metrics":[19],"like":[20,28],"faithfulness":[21],"and":[22,24,48,82,98,138],"compactness,":[23],"human-grounded":[26],"proxies":[27],"subjective":[29],"ratings":[30],"or":[31],"prediction":[32],"accuracy.":[33],"We":[34,55],"suggest":[35],"evaluating":[36],"XRL":[37,71,95,114,126,165,181],"methods":[38,72,166,182],"by":[39],"how":[40],"effectively":[41],"their":[42],"generated":[43],"explanations":[44],"help":[45],"to":[46,73,107],"diagnose":[47,74],"fix":[49],"malfunctioning":[50],"reinforcement":[51],"learning":[52],"(RL)":[53],"agents.":[54],"propose":[56,174],"EvalXRL,":[57],"in":[60,78,183],"which":[61],"Large":[63],"Language":[64],"Model":[65],"(LLM)":[66],"coding":[67,117],"agent":[68,118],"uses":[69,99],"different":[70],"held-out":[76],"malfunction":[77,93],"an":[79],"RL":[80,105],"agent,":[81],"then":[83],"repair":[84],"it.":[85],"Our":[86],"proposed":[87],"iterates":[89],"across":[90],"(environment":[91],"$\\times$":[92,94],"method)":[96],"tuples":[97],"the":[100,104,121,125,140,161,175],"reward":[101],"signal":[102],"of":[103,160,179],"agents":[106],"form":[108,131],"final":[110],"score":[111],"each":[113],"method.":[115,163],"The":[116],"may":[119,153],"use":[120],"method":[122,141],"interactively:":[123],"invoke":[124,139],"method,":[127],"process":[128],"its":[129],"output,":[130],"new":[132],"hypotheses":[133],"what":[135],"is":[136],"broken,":[137],"again":[142],"with":[143],"parameters":[144],"adjusted":[145],"testing":[147],"these":[148],"hypotheses.":[149],"closed-loop":[151,184],"structure":[152],"be":[154],"described":[155],"as":[156],"simplified":[158],"version":[159],"scientific":[162],"Some":[164],"provide":[167],"self-evaluations":[168],"that":[169],"follow":[170],"this":[171],"pattern;":[172],"we":[173],"first":[176],"head-to-head":[177],"comparison":[178],"multiple":[180],"usage.":[185]},"counts_by_year":[],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2026-08-20T00:00:00"}