{"id":"https://openalex.org/W7225006381","doi":"https://doi.org/10.5281/zenodo.23846266","title":"No Step Looked Wrong: FadeBench, a Benchmark for Incremental Moral Drift in Small Language Models","display_name":"No Step Looked Wrong: FadeBench, a Benchmark for Incremental Moral Drift in Small Language Models","publication_year":2026,"publication_date":"2026-10-31","ids":{"openalex":"https://openalex.org/W7225006381","doi":"https://doi.org/10.5281/zenodo.23846266"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.23846266","doi":"https://doi.org/10.5281/zenodo.23846266","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23846266","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},"volume":null,"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.23846266","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5149611876","display_name":"Becky Northaven","orcid":"https://orcid.org/0009-0007-6573-6544","observed_orcids":["https://orcid.org/0009-0007-6573-6544"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Northaven, Becky R.","raw_affiliation_strings":["Sequim Electronics"],"raw_orcid":"https://orcid.org/0009-0007-6573-6544","affiliations":[{"raw_affiliation_string":"Sequim Electronics","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5149611876"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"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/T10883","display_name":"Ethics and Social Impacts of AI","score":0.9096999764442444,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T10883","display_name":"Ethics and Social Impacts of AI","score":0.9096999764442444,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T11689","display_name":"Adversarial Robustness in Machine Learning","score":0.050700001418590546,"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/T11058","display_name":"Ethics in Business and Education","score":0.005499999970197678,"subfield":{"id":"https://openalex.org/subfields/1802","display_name":"Information Systems and Management"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/small-language-models","display_name":"small language models","score":0.8220000267028809},{"id":"https://openalex.org/keywords/ethical-drift","display_name":"ethical drift","score":0.3479999899864197},{"id":"https://openalex.org/keywords/incremental-decision-making","display_name":"incremental decision-making","score":0.5019999742507935},{"id":"https://openalex.org/keywords/moral-reasoning","display_name":"moral reasoning","score":0.36399999260902405},{"id":"https://openalex.org/keywords/ethical-dilemmas","display_name":"ethical dilemmas","score":0.41499999165534973},{"id":"https://openalex.org/keywords/harm-to-others","display_name":"harm to others","score":0.32100000977516174},{"id":"https://openalex.org/keywords/large-language-model-safety","display_name":"large language model safety","score":0.5329999923706055}],"concepts":[{"id":"https://openalex.org/C2778689934","wikidata":"https://www.wikidata.org/wiki/Q1313396","display_name":"Headline","level":2,"score":0.616599977016449},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5788000226020813},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5733000040054321},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.5202999711036682},{"id":"https://openalex.org/C2777267654","wikidata":"https://www.wikidata.org/wiki/Q3519023","display_name":"Test (biology)","level":2,"score":0.4984999895095825},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.498199999332428},{"id":"https://openalex.org/C2779545769","wikidata":"https://www.wikidata.org/wiki/Q5135364","display_name":"Closeness","level":2,"score":0.42980000376701355},{"id":"https://openalex.org/C42058472","wikidata":"https://www.wikidata.org/wiki/Q810214","display_name":"Base (topology)","level":2,"score":0.3978999853134155},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.396699994802475},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.3944999873638153},{"id":"https://openalex.org/C2982912361","wikidata":"https://www.wikidata.org/wiki/Q1851867","display_name":"Mental model","level":2,"score":0.39320001006126404},{"id":"https://openalex.org/C534859617","wikidata":"https://www.wikidata.org/wiki/Q5434357","display_name":"Fantasy","level":2,"score":0.38679999113082886},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.38370001316070557},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.3788999915122986},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.33469998836517334},{"id":"https://openalex.org/C2781035248","wikidata":"https://www.wikidata.org/wiki/Q186150","display_name":"Fallacy","level":2,"score":0.3296999931335449},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.3118000030517578},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.30869999527931213},{"id":"https://openalex.org/C2780876879","wikidata":"https://www.wikidata.org/wiki/Q3054749","display_name":"Meaning (existential)","level":2,"score":0.30219998955726624},{"id":"https://openalex.org/C41065033","wikidata":"https://www.wikidata.org/wiki/Q2825412","display_name":"Adversary","level":2,"score":0.3012999892234802},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.3001999855041504},{"id":"https://openalex.org/C77805123","wikidata":"https://www.wikidata.org/wiki/Q161272","display_name":"Social psychology","level":1,"score":0.2964000105857849},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.29490000009536743},{"id":"https://openalex.org/C44154836","wikidata":"https://www.wikidata.org/wiki/Q45045","display_name":"Simulation","level":1,"score":0.27059999108314514},{"id":"https://openalex.org/C80519477","wikidata":"https://www.wikidata.org/wiki/Q3532236","display_name":"Scenario testing","level":3,"score":0.26429998874664307},{"id":"https://openalex.org/C42475967","wikidata":"https://www.wikidata.org/wiki/Q194292","display_name":"Operations research","level":1,"score":0.25189998745918274},{"id":"https://openalex.org/C78821406","wikidata":"https://www.wikidata.org/wiki/Q391810","display_name":"Think aloud protocol","level":3,"score":0.2508000135421753}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.23846266","doi":"https://doi.org/10.5281/zenodo.23846266","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23846266","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},{"id":"doi:10.5281/zenodo.23846267","doi":"https://doi.org/10.5281/zenodo.23846267","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23846267","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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":"doi:10.5281/zenodo.23846266","doi":"https://doi.org/10.5281/zenodo.23846266","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.23846266","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"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"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":1,"referenced_works":["https://openalex.org/W7207612778"],"related_works":["https://openalex.org/W7164388337","https://openalex.org/W7117322593","https://openalex.org/W7221816594","https://openalex.org/W7138858787","https://openalex.org/W7154419022","https://openalex.org/W7148358204","https://openalex.org/W7224491625","https://openalex.org/W7223017469","https://openalex.org/W7148324062","https://openalex.org/W4311553886"],"abstract_inverted_index":{"At":[0],"the":[1,28,38,118,141,169,176,194,197,219,227,234,237,266,271,279,299,307,317,355,359,362,386,396,399],"end":[2,235,397],"of":[3,6,105,138,204,222,236,253,260,298,336,340,344,369,398],"a":[4,9,43,61,103,111,115,146,150,155,161,186,223,251,257,314,321,342],"run":[5,297,403],"small":[7,10,80,181],"steps,":[8],"language":[11,39,418],"model":[12,40,62,69,221,368,419],"can":[13,70],"take":[14],"an":[15,295],"act":[16],"it":[17,148,281,333],"usually":[18],"declines":[19],"when":[20,265],"asked":[21],"outright.":[22],"Each":[23,101],"step":[24],"looks":[25],"defensible":[26],"against":[27,286,338],"one":[29,53,56,367],"before":[30],"it,":[31,213],"and":[32,74,160,202,239,376,383,415],"no":[33,416],"user":[34],"or":[35,58,92],"adversary":[36],"pushes":[37],"actor":[41],"toward":[42],"bad":[44],"choice.":[45],"This":[46],"matters":[47],"because":[48],"many":[49],"safety":[50],"batteries":[51],"judge":[52],"answer":[54,354],"to":[55,140,262,348,407],"prompt,":[57],"test":[59,192],"how":[60],"holds":[63,171],"up":[64],"under":[65],"sustained":[66],"pressure.":[67],"A":[68,164],"pass":[71],"those":[72,330],"tests":[73],"still":[75],"drift":[76,228],"as":[77,277],"its":[78,129,143,152,216,380],"own":[79,130,144,153],"choices":[81],"create":[82],"surprising,":[83],"harmful":[84],"consequences":[85],"over":[86,325],"turns.":[87],"FadeBench":[88],"measures":[89],"this":[90,303],"ethical":[91],"moral":[93],"drift.":[94],"The":[95,120,132,190,401],"battery":[96],"has":[97,174],"38":[98],"scenario":[99,208,249],"items.":[100],"is":[102,124,282,313,334,364,412],"ladder":[104,238],"four":[106,180,352],"decisions,":[107],"called":[108],"rungs,":[109],"in":[110,128,210,231,244,275,284,288,306],"fantasy":[112],"scene":[113,411],"where":[114],"dependent":[116],"bears":[117],"cost.":[119],"same":[121,300,318,356,360],"decision":[122],"point":[123],"also":[125],"put":[126],"cold,":[127,246],"words.":[131],"dependents":[133],"sit":[134],"on":[135,185,302,316,366,373,385],"seven":[136],"rings":[137],"closeness":[139],"actor:":[142],"self,":[145],"person":[147],"knows,":[149],"stranger,":[151],"pet,":[154],"neighborhood":[156],"animal,":[157],"wild":[158,162],"prey":[159],"predator.":[163],"further":[165],"place,":[166],"apart":[167],"from":[168,294,404],"seven,":[170],"someone":[172],"who":[173],"wronged":[175],"actor.":[177],"We":[178],"ran":[179],"open":[182],"models":[183,353,402],"bare,":[184],"plain":[187],"inference":[188],"server.":[189],"headline":[191],"pools":[193],"baseline":[195],"set,":[196],"battery's":[198],"first":[199],"14":[200],"items":[201,301,357],"three":[203],"their":[205],"variants,":[206],"17":[207],"versions":[209],"all.":[211],"On":[212,270],"gemma-4-E4B":[214,377],"with":[215,256,379],"reasoning":[217],"on,":[218],"base":[220,304],"deployed":[224],"system,":[225],"takes":[226],"15":[229],"times":[230,243],"100":[232,245,285],"at":[233,358,395],"fewer":[240],"than":[241,394],"2":[242,287],"counting":[247],"each":[248],"once:":[250],"gap":[252,343,363],"13.2":[254],"points,":[255],"95%":[258],"interval":[259],"2.6":[261],"25.9":[263],"points":[264,346],"scenarios":[267,273],"are":[268],"resampled.":[269],"five":[272,291,331],"named":[274],"advance":[276],"carrying":[278],"effect,":[280],"43":[283],"99.":[289],"Those":[290],"were":[292],"chosen":[293],"earlier":[296],"model,":[305,319],"deployed-system":[308],"configurations,":[309],"so":[310],"that":[311],"figure":[312],"replication":[315],"not":[320],"held-out":[322],"prediction.":[323],"Pooled":[324],"all":[326,351],"samples,":[327],"which":[328,389],"counts":[329],"twice,":[332],"47":[335],"220":[337],"4":[339],"219,":[341],"19.5":[345],"(4.0":[347],"34.4).":[349],"Where":[350],"settings,":[361],"clear":[365],"four,":[370],"qwen3-8b,":[371],"uncertain":[372],"two,":[374],"phi-4-mini":[375],"itself":[378],"thinking":[381],"off,":[382],"absent":[384],"fourth,":[387],"mistral-nemo-12b,":[388],"drifts":[390],"more":[391],"often":[392],"cold":[393],"ladder.":[400],"roughly":[405],"4B":[406],"12B":[408],"parameters,":[409],"every":[410],"low-key":[413],"fantasy,":[414],"frontier":[417],"was":[420],"tested.":[421]},"counts_by_year":[],"updated_date":"2026-11-02T09:45:03.842371","created_date":"2026-11-01T00:00:00"}