{"id":"https://openalex.org/W7154626005","doi":"https://doi.org/10.48550/arxiv.2604.13075","title":"DeEscalWild: A Real-World Benchmark for Automated De-Escalation Training with SLMs","display_name":"DeEscalWild: A Real-World Benchmark for Automated De-Escalation Training with SLMs","publication_year":2026,"publication_date":"2026-03-20","ids":{"openalex":"https://openalex.org/W7154626005","doi":"https://doi.org/10.48550/arxiv.2604.13075"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2604.13075","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.13075","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2604.13075","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5111301681","display_name":"Md Hasebul Hasan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hasan, Md Hasebul","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5092667135","display_name":"Krity Haque Charu","orcid":"https://orcid.org/0009-0008-1336-5371"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Charu, Krity Haque","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5093069317","display_name":"Eshwara Prasad Sridhar","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sridhar, Eshwara Prasad","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5081413226","display_name":"Shuchisnigdha Deb","orcid":"https://orcid.org/0000-0001-5626-2825"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Deb, Shuchisnigdha","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5133777758","display_name":"Mohammad A. Islam","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Islam, Mohammad A.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.25609999895095825,"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":0.25609999895095825,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.06689999997615814,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.04899999871850014,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.6822999715805054},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.631600022315979},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.631600022315979},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.5382999777793884},{"id":"https://openalex.org/keywords/field","display_name":"Field (mathematics)","score":0.5185999870300293},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.5113999843597412},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.42980000376701355},{"id":"https://openalex.org/keywords/bridge","display_name":"Bridge (graph theory)","score":0.4269999861717224},{"id":"https://openalex.org/keywords/footprint","display_name":"Footprint","score":0.35280001163482666}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7757999897003174},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6822999715805054},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.631600022315979},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.631600022315979},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.5382999777793884},{"id":"https://openalex.org/C9652623","wikidata":"https://www.wikidata.org/wiki/Q190109","display_name":"Field (mathematics)","level":2,"score":0.5185999870300293},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.5113999843597412},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.45809999108314514},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.44929999113082886},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.42980000376701355},{"id":"https://openalex.org/C100776233","wikidata":"https://www.wikidata.org/wiki/Q2532492","display_name":"Bridge (graph theory)","level":2,"score":0.4269999861717224},{"id":"https://openalex.org/C132943942","wikidata":"https://www.wikidata.org/wiki/Q2562511","display_name":"Footprint","level":2,"score":0.35280001163482666},{"id":"https://openalex.org/C2777655017","wikidata":"https://www.wikidata.org/wiki/Q1501161","display_name":"Toolbox","level":2,"score":0.3488999903202057},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.3481999933719635},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.32359999418258667},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.32170000672340393},{"id":"https://openalex.org/C105002631","wikidata":"https://www.wikidata.org/wiki/Q4833645","display_name":"Subject-matter expert","level":3,"score":0.31690001487731934},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.314300000667572},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.2922999858856201},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.2831000089645386},{"id":"https://openalex.org/C74912251","wikidata":"https://www.wikidata.org/wiki/Q6815727","display_name":"Memory footprint","level":2,"score":0.27730000019073486},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.273499995470047},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.26649999618530273},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.2596000134944916},{"id":"https://openalex.org/C75684735","wikidata":"https://www.wikidata.org/wiki/Q858810","display_name":"Big data","level":2,"score":0.2574999928474426},{"id":"https://openalex.org/C186644900","wikidata":"https://www.wikidata.org/wiki/Q194152","display_name":"Parsing","level":2,"score":0.25290000438690186},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.2524000108242035},{"id":"https://openalex.org/C49895821","wikidata":"https://www.wikidata.org/wiki/Q5227368","display_name":"Data verification","level":2,"score":0.2522999942302704},{"id":"https://openalex.org/C175309249","wikidata":"https://www.wikidata.org/wiki/Q725864","display_name":"Pipeline transport","level":2,"score":0.25040000677108765}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2604.13075","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.13075","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2604.13075","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2604.13075","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,"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"},"sustainable_development_goals":[{"score":0.5583221316337585,"display_name":"Industry, innovation and infrastructure","id":"https://metadata.un.org/sdg/9"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Effective":[0],"de-escalation":[1],"is":[2],"critical":[3,60],"for":[4,35,43,192],"law":[5],"enforcement":[6],"safety":[7],"and":[8,17,149,195,208],"community":[9],"trust,":[10],"yet":[11],"traditional":[12],"training":[13,65,198],"methods":[14],"lack":[15],"scalability":[16],"realism.":[18],"While":[19],"Large":[20],"Language":[21,48],"Models":[22,49],"(LLMs)":[23],"enable":[24],"dynamic,":[25],"open-ended":[26],"simulations,":[27],"their":[28,139],"substantial":[29],"computational":[30,184],"footprint":[31],"renders":[32],"them":[33],"impractical":[34],"deployment":[36],"on":[37,134],"the":[38,160,183,189,201],"lightweight,":[39],"portable":[40],"hardware":[41],"required":[42],"immersive":[44],"field":[45],"training.":[46],"Small":[47],"(SLMs)":[50],"offer":[51],"a":[52,59,74,80,100,180],"viable":[53],"real-time":[54],"alternative":[55],"but":[56],"suffer":[57],"from":[58,79,88],"scarcity":[61],"of":[62,83,182],"high-quality,":[63],"domain-specific":[64],"data.":[66],"To":[67],"bridge":[68],"this":[69,135],"gap,":[70],"we":[71,98],"present":[72],"DeEscalWild,":[73],"novel":[75],"benchmark":[76],"dataset":[77],"curated":[78],"multi-stage":[81],"pipeline":[82],"in-the-wild":[84],"police-civilian":[85],"interactions":[86],"extracted":[87],"publicly":[89,204],"available":[90],"video":[91],"repositories.":[92],"Starting":[93],"with":[94,108,179],"5,000":[95],"raw":[96],"inputs,":[97],"employed":[99],"rigorous":[101],"hybrid":[102],"filtering":[103],"process":[104],"combining":[105],"human-in-the-loop":[106],"verification":[107],"LLM-as-a-Judge":[109],"evaluation":[110,151],"to":[111],"distill":[112],"1,500":[113],"high-fidelity":[114],"scenarios.":[115],"The":[116],"resulting":[117],"corpus":[118],"comprises":[119],"285,887":[120],"dialogue":[121],"turns,":[122],"totaling":[123],"approximately":[124],"4.7":[125],"million":[126],"tokens.":[127],"Extensive":[128],"experiments":[129],"demonstrate":[130],"that":[131,172],"SLMs":[132,174],"fine-tuned":[133,155],"data":[136],"significantly":[137],"outperform":[138],"base":[140],"counterparts":[141],"across":[142],"ROUGE-L,":[143],"BLEU-4,":[144],"METEOR,":[145],"BERTScore,":[146],"Realism":[147],"Score,":[148],"human":[150],"metrics.":[152],"Notably,":[153],"our":[154,206],"Qwen":[156],"2.5":[157,163],"(3B-Instruct)":[158],"surpasses":[159],"general-purpose":[161],"Gemini":[162],"Flash":[164],"model":[165],"when":[166],"evaluated":[167],"under":[168],"equivalent":[169],"conditions,":[170],"demonstrating":[171],"domain-optimized":[173],"can":[175],"achieve":[176],"superior":[177],"performance":[178],"fraction":[181],"cost.":[185],"This":[186],"work":[187],"establishes":[188],"foundational":[190],"infrastructure":[191],"accessible,":[193],"low-latency,":[194],"privacy-preserving":[196],"officer":[197],"systems":[199],"at":[200],"edge.":[202],"We":[203],"release":[205],"code(https://github.com/Hasebul/DeEscalWild-Benchmark-Framework)":[207],"dataset(https://doi.org/10.7910/DVN/CWMCZI).":[209]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-04-17T00:00:00"}