{"id":"https://openalex.org/W7225614095","doi":"https://doi.org/10.1038/s41598-026-78750-5","title":"Diagnostic probes reveal shortcuts induced by the benchmark protocol in vision-language model annotation pipelines for video-based self-harm detection","display_name":"Diagnostic probes reveal shortcuts induced by the benchmark protocol in vision-language model annotation pipelines for video-based self-harm detection","publication_year":2026,"publication_date":"2026-11-09","ids":{"openalex":"https://openalex.org/W7225614095","doi":"https://doi.org/10.1038/s41598-026-78750-5"},"language":"en","primary_location":{"id":"doi:10.1038/s41598-026-78750-5","doi":"https://doi.org/10.1038/s41598-026-78750-5","is_oa":true,"landing_page_url":"https://doi.org/10.1038/s41598-026-78750-5","pdf_url":"https://www.nature.com/articles/s41598-026-78750-5_reference.pdf","source":{"id":"https://openalex.org/S196734849","display_name":"Scientific Reports","issn_l":"2045-2322","issn":["2045-2322"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["anvur-class-a","anvur-scientific","cwts-core","doaj","dongbi-c","doyens","jufo-1","ki-jl-1","medline","norway-1","poland-140","russia-white-list-1","vabb-shw"],"host_organization":"https://openalex.org/P4310319908","host_organization_name":"Nature Portfolio","host_organization_lineage":["https://openalex.org/P4310319908","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Nature Portfolio","Springer Nature"],"type":"journal"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Scientific Reports","raw_type":"journal-article"},"volume":null,"type":"article","is_traditional":true,"indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://www.nature.com/articles/s41598-026-78750-5_reference.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5158817056","display_name":"Yourack Lee","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4405258752","display_name":"Korea University Guro Hospital","ror":"https://ror.org/0154bb690","country_code":"KR","type":"healthcare","lineage":["https://openalex.org/I197347611","https://openalex.org/I2799980853","https://openalex.org/I4405258752"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Yourack Lee","raw_affiliation_strings":["Biomedical Research Center, Korea University Guro Hospital, Seoul, 08308, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Biomedical Research Center, Korea University Guro Hospital, Seoul, 08308, Republic of Korea","institution_ids":["https://openalex.org/I4405258752"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008377585","display_name":"Hyun\u2010Ghang Jeong","orcid":"https://orcid.org/0000-0002-0318-5069","observed_orcids":["https://orcid.org/0000-0002-0318-5069"]},"institutions":[{"id":"https://openalex.org/I4405258752","display_name":"Korea University Guro Hospital","ror":"https://ror.org/0154bb690","country_code":"KR","type":"healthcare","lineage":["https://openalex.org/I197347611","https://openalex.org/I2799980853","https://openalex.org/I4405258752"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Hyun-Ghang Jeong","raw_affiliation_strings":["Department of Psychiatry, Korea University Guro Hospital, Korea University College of Medicine, Seoul, 08308, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Psychiatry, Korea University Guro Hospital, Korea University College of Medicine, Seoul, 08308, Republic of Korea","institution_ids":["https://openalex.org/I4405258752"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5124464268","display_name":"Hyun-Soo Choi","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I118373667","display_name":"Seoul National University of Science and Technology","ror":"https://ror.org/00chfja07","country_code":"KR","type":"education","lineage":["https://openalex.org/I118373667"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Hyun-Soo Choi","raw_affiliation_strings":["Department of Computer Science and Engineering, Seoul National University of Science and Technology, Seoul, 01811, Republic of Korea"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, Seoul National University of Science and Technology, Seoul, 01811, Republic of Korea","institution_ids":["https://openalex.org/I118373667"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":["https://openalex.org/A5008377585","https://openalex.org/A5124464268"],"corresponding_institution_ids":["https://openalex.org/I118373667","https://openalex.org/I4405258752"],"apc_list":{"value":2690,"currency":"USD","value_usd":2690},"apc_paid":{"value":2690,"currency":"USD","value_usd":2690},"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.0,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.5386000275611877,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.5386000275611877,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T12488","display_name":"Mental Health via Writing","score":0.1256999969482422,"subfield":{"id":"https://openalex.org/subfields/3207","display_name":"Social Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T10376","display_name":"Suicide and Self-Harm Studies","score":0.03880000114440918,"subfield":{"id":"https://openalex.org/subfields/3203","display_name":"Clinical Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/vision-language-models","display_name":"vision-language models","score":0.9319999814033508},{"id":"https://openalex.org/keywords/self-harm-detection","display_name":"self-harm detection","score":0.7860000133514404},{"id":"https://openalex.org/keywords/video-annotation","display_name":"video annotation","score":0.5870000123977661},{"id":"https://openalex.org/keywords/diagnostic-probes","display_name":"diagnostic probes","score":0.8149999976158142},{"id":"https://openalex.org/keywords/frame-selection","display_name":"frame selection","score":0.4729999899864197},{"id":"https://openalex.org/keywords/prompt-templates","display_name":"prompt templates","score":0.4440000057220459}],"concepts":[{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.8327999711036682},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7347000241279602},{"id":"https://openalex.org/C97931131","wikidata":"https://www.wikidata.org/wiki/Q5282087","display_name":"Discriminative model","level":2,"score":0.7290999889373779},{"id":"https://openalex.org/C95623464","wikidata":"https://www.wikidata.org/wiki/Q1096149","display_name":"Classifier (UML)","level":2,"score":0.6216999888420105},{"id":"https://openalex.org/C2776321320","wikidata":"https://www.wikidata.org/wiki/Q857525","display_name":"Annotation","level":2,"score":0.567799985408783},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5461999773979187},{"id":"https://openalex.org/C175309249","wikidata":"https://www.wikidata.org/wiki/Q725864","display_name":"Pipeline transport","level":2,"score":0.5414999723434448},{"id":"https://openalex.org/C2780385302","wikidata":"https://www.wikidata.org/wiki/Q367158","display_name":"Protocol (science)","level":3,"score":0.4717000126838684},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.41370001435279846},{"id":"https://openalex.org/C148524875","wikidata":"https://www.wikidata.org/wiki/Q6975395","display_name":"F1 score","level":2,"score":0.4059000015258789},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.3905999958515167},{"id":"https://openalex.org/C150921843","wikidata":"https://www.wikidata.org/wiki/Q1170431","display_name":"Resampling","level":2,"score":0.3788999915122986},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3564999997615814},{"id":"https://openalex.org/C21308566","wikidata":"https://www.wikidata.org/wiki/Q7169365","display_name":"Permutation (music)","level":2,"score":0.3181000053882599},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.3167000114917755},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.3122999966144562},{"id":"https://openalex.org/C169258074","wikidata":"https://www.wikidata.org/wiki/Q245748","display_name":"Random forest","level":2,"score":0.2849999964237213},{"id":"https://openalex.org/C16910744","wikidata":"https://www.wikidata.org/wiki/Q7705759","display_name":"Test data","level":2,"score":0.2847000062465668},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.27900001406669617},{"id":"https://openalex.org/C126042441","wikidata":"https://www.wikidata.org/wiki/Q1324888","display_name":"Frame (networking)","level":2,"score":0.2703000009059906},{"id":"https://openalex.org/C2778739407","wikidata":"https://www.wikidata.org/wiki/Q165372","display_name":"CLIPS","level":2,"score":0.2669000029563904},{"id":"https://openalex.org/C2778775528","wikidata":"https://www.wikidata.org/wiki/Q5135432","display_name":"Closing (real estate)","level":2,"score":0.257099986076355},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.25519999861717224}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1038/s41598-026-78750-5","doi":"https://doi.org/10.1038/s41598-026-78750-5","is_oa":true,"landing_page_url":"https://doi.org/10.1038/s41598-026-78750-5","pdf_url":"https://www.nature.com/articles/s41598-026-78750-5_reference.pdf","source":{"id":"https://openalex.org/S196734849","display_name":"Scientific Reports","issn_l":"2045-2322","issn":["2045-2322"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["anvur-class-a","anvur-scientific","cwts-core","doaj","dongbi-c","doyens","jufo-1","ki-jl-1","medline","norway-1","poland-140","russia-white-list-1","vabb-shw"],"host_organization":"https://openalex.org/P4310319908","host_organization_name":"Nature Portfolio","host_organization_lineage":["https://openalex.org/P4310319908","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Nature Portfolio","Springer Nature"],"type":"journal"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Scientific Reports","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1038/s41598-026-78750-5","doi":"https://doi.org/10.1038/s41598-026-78750-5","is_oa":true,"landing_page_url":"https://doi.org/10.1038/s41598-026-78750-5","pdf_url":"https://www.nature.com/articles/s41598-026-78750-5_reference.pdf","source":{"id":"https://openalex.org/S196734849","display_name":"Scientific Reports","issn_l":"2045-2322","issn":["2045-2322"],"is_oa":true,"is_in_doaj":true,"is_core":true,"listed_in":["anvur-class-a","anvur-scientific","cwts-core","doaj","dongbi-c","doyens","jufo-1","ki-jl-1","medline","norway-1","poland-140","russia-white-list-1","vabb-shw"],"host_organization":"https://openalex.org/P4310319908","host_organization_name":"Nature Portfolio","host_organization_lineage":["https://openalex.org/P4310319908","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Nature Portfolio","Springer Nature"],"type":"journal"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Scientific Reports","raw_type":"journal-article"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320322107","display_name":"Korea Health Industry Development Institute","ror":"https://ror.org/00fdzyk40"},{"id":"https://openalex.org/F4320322120","display_name":"National Research Foundation of Korea","ror":"https://ror.org/013aysd81"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7225614095.pdf","grobid_xml":"https://content.openalex.org/works/W7225614095.grobid-xml"},"referenced_works_count":4,"referenced_works":["https://openalex.org/W4387724705","https://openalex.org/W4414364960","https://openalex.org/W7133301267","https://openalex.org/W7222265838"],"related_works":["https://openalex.org/W4417141639","https://openalex.org/W7221521445","https://openalex.org/W7216135133","https://openalex.org/W7211999815","https://openalex.org/W4402670365","https://openalex.org/W4415800385","https://openalex.org/W7140316077","https://openalex.org/W7160869148","https://openalex.org/W7221074449","https://openalex.org/W7223012147"],"abstract_inverted_index":{"Vision-language":[0],"models":[1],"(VLMs)":[2],"are":[3,192],"increasingly":[4],"used":[5],"as":[6,29],"annotators:":[7],"a":[8,15,43,80,159,163,170,176],"model":[9],"captions":[10],"clinical":[11,32,194],"images":[12],"or":[13],"video,":[14],"classifier":[16],"is":[17,27],"trained":[18],"on":[19,46,175],"the":[20,23,47,52,88,92,116,132,140],"captions,":[21],"and":[22,101,134,139,155],"resulting":[24],"benchmark":[25,69,89],"score":[26,44,171],"read":[28],"evidence":[30],"of":[31,99,136,181],"capability.":[33],"We":[34],"describe":[35],"two":[36],"diagnostic":[37],"probes":[38,167],"that":[39,109,172],"test":[40],"whether":[41],"such":[42],"depends":[45],"target":[48],"finding.":[49],"Necessity":[50],"deletes":[51],"most":[53,94],"class-discriminative":[54],"caption":[55],"tokens;":[56],"sufficiency":[57],"classifies":[58],"from":[59,115,122,151],"content-agnostic":[60],"surface-form":[61,103,111],"statistics":[62,104],"alone.":[63],"On":[64],"an":[65],"acted":[66],"self-harm":[67],"video":[68],"(5,183":[70],"clips;":[71],"seven":[72],"actors;":[73],"leave-one-subject-out":[74],"evaluation),":[75],"four":[76],"VLM":[77],"pipelines":[78,157,183],"with":[79,127],"clinically":[81],"engineered":[82],"prompt":[83,141],"reached":[84],"F1":[85],"0.784\u20130.944":[86],"under":[87],"protocol.":[90],"Deleting":[91],"40":[93],"discriminative":[95],"tokens":[96],"retained":[97,158],"92\u2013100%":[98],"F1,":[100],"eight":[102],"recovered":[105],"83\u201395%.":[106],"Re-analysis":[107],"showed":[108],"this":[110,188],"signal":[112,161],"arose":[113],"mainly":[114],"protocol,":[117],"which":[118],"trims":[119],"boundary":[120],"segments":[121],"normal":[123],"clips":[124],"only":[125],"and,":[126],"our":[128],"captioning":[129],"order,":[130],"changes":[131],"number":[133],"position":[135],"captioned":[137],"frames,":[138],"template,":[142],"by":[143],"class.":[144],"With":[145],"label-independent":[146],"frame":[147],"selection,":[148],"AUC":[149],"fell":[150],"0.89\u2013\u200b0.99":[152],"to":[153],"0.53\u20130.65,":[154],"three":[156],"modest":[160],"above":[162],"permutation":[164],"null.":[165],"The":[166],"thus":[168],"flagged":[169],"rested":[173],"largely":[174],"protocol":[177],"artifact.":[178],"Benchmark":[179],"scores":[180],"VLM-annotator":[182],"should":[184],"be":[185],"checked":[186],"in":[187],"way":[189],"before":[190],"they":[191],"given":[193],"weight.":[195]},"counts_by_year":[],"updated_date":"2026-11-11T12:16:26.678958","created_date":"2026-11-10T00:00:00"}