{"id":"https://openalex.org/W4391591738","doi":"https://doi.org/10.48550/arxiv.2402.01822","title":"Building Guardrails for Large Language Models","display_name":"Building Guardrails for Large Language Models","publication_year":2024,"publication_date":"2024-02-02","ids":{"openalex":"https://openalex.org/W4391591738","doi":"https://doi.org/10.48550/arxiv.2402.01822"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2402.01822","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.01822","pdf_url":"https://arxiv.org/pdf/2402.01822","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","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/2402.01822","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5100650289","display_name":"Yi Dong","orcid":"https://orcid.org/0000-0003-3047-7777","observed_orcids":["https://orcid.org/0000-0003-3047-7777"]},"institutions":[{"id":"https://openalex.org/I146655781","display_name":"University of Liverpool","ror":"https://ror.org/04xs57h96","country_code":"GB","type":"education","lineage":["https://openalex.org/I146655781"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Dong, Yi","raw_affiliation_strings":["Department of Computer Science , Uni- versity of Liverpool , UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Uni- versity of Liverpool , UK","institution_ids":["https://openalex.org/I146655781"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054272337","display_name":"Ronghui Mu","orcid":"https://orcid.org/0000-0001-6150-4948","observed_orcids":["https://orcid.org/0000-0001-6150-4948"]},"institutions":[{"id":"https://openalex.org/I146655781","display_name":"University of Liverpool","ror":"https://ror.org/04xs57h96","country_code":"GB","type":"education","lineage":["https://openalex.org/I146655781"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Mu, Ronghui","raw_affiliation_strings":["Department of Computer Science , Uni- versity of Liverpool , UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Uni- versity of Liverpool , UK","institution_ids":["https://openalex.org/I146655781"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048979254","display_name":"Gaojie Jin","orcid":"https://orcid.org/0000-0003-0240-3033","observed_orcids":["https://orcid.org/0000-0003-0240-3033"]},"institutions":[{"id":"https://openalex.org/I19820366","display_name":"Chinese Academy of Sciences","ror":"https://ror.org/034t30j35","country_code":"CN","type":"government","lineage":["https://openalex.org/I19820366"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jin, Gaojie","raw_affiliation_strings":["Key Laboratory of System Software (Chinese Academy of Sciences) and"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Key Laboratory of System Software (Chinese Academy of Sciences) and","institution_ids":["https://openalex.org/I19820366"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5107746894","display_name":"Yi Qi","orcid":"https://orcid.org/0000-0002-8758-930X","observed_orcids":["https://orcid.org/0000-0002-8758-930X"]},"institutions":[{"id":"https://openalex.org/I146655781","display_name":"University of Liverpool","ror":"https://ror.org/04xs57h96","country_code":"GB","type":"education","lineage":["https://openalex.org/I146655781"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Qi, Yi","raw_affiliation_strings":["Department of Computer Science , Uni- versity of Liverpool , UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Uni- versity of Liverpool , UK","institution_ids":["https://openalex.org/I146655781"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101242822","display_name":"Jinwei Hu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I146655781","display_name":"University of Liverpool","ror":"https://ror.org/04xs57h96","country_code":"GB","type":"education","lineage":["https://openalex.org/I146655781"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Hu, Jinwei","raw_affiliation_strings":["Department of Computer Science , Uni- versity of Liverpool , UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Uni- versity of Liverpool , UK","institution_ids":["https://openalex.org/I146655781"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100635681","display_name":"Xingyu Zhao","orcid":"https://orcid.org/0000-0002-3474-349X","observed_orcids":["https://orcid.org/0000-0002-3474-349X","https://orcid.org/0009-0005-9422-042X"]},"institutions":[{"id":"https://openalex.org/I39555362","display_name":"University of Warwick","ror":"https://ror.org/01a77tt86","country_code":"GB","type":"education","lineage":["https://openalex.org/I39555362"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Zhao, Xingyu","raw_affiliation_strings":["WMG , University of Warwick , Warwick , UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"WMG , University of Warwick , Warwick , UK","institution_ids":["https://openalex.org/I39555362"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Meng, Jie","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Meng, Jie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5074225885","display_name":"Wenjie Ruan","orcid":"https://orcid.org/0000-0002-8311-8738","observed_orcids":["https://orcid.org/0000-0002-8311-8738"]},"institutions":[{"id":"https://openalex.org/I146655781","display_name":"University of Liverpool","ror":"https://ror.org/04xs57h96","country_code":"GB","type":"education","lineage":["https://openalex.org/I146655781"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Ruan, Wenjie","raw_affiliation_strings":["Department of Computer Science , Uni- versity of Liverpool , UK"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Uni- versity of Liverpool , UK","institution_ids":["https://openalex.org/I146655781"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5121195030","display_name":"Xiaowei Huang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I146655781","display_name":"University of Liverpool","ror":"https://ror.org/04xs57h96","country_code":"GB","type":"education","lineage":["https://openalex.org/I146655781"]},{"id":"https://openalex.org/I19820366","display_name":"Chinese Academy of Sciences","ror":"https://ror.org/034t30j35","country_code":"CN","type":"government","lineage":["https://openalex.org/I19820366"]},{"id":"https://openalex.org/I4210128818","display_name":"Institute of Software","ror":"https://ror.org/033dfsn42","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210128818"]},{"id":"https://openalex.org/I4391767820","display_name":"State Key Laboratory of Computer Science","ror":"https://ror.org/01hsx4r68","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210128818","https://openalex.org/I4391767820"]}],"countries":["CN","GB"],"is_corresponding":false,"raw_author_name":"Huang, Xiaowei","raw_affiliation_strings":["Department of Computer Science , Uni- versity of Liverpool , UK","State Key Laboratory of Com- puter Science , Institute of Software , Chinese Academy of Sciences"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Department of Computer Science , Uni- versity of Liverpool , UK","institution_ids":["https://openalex.org/I146655781"]},{"raw_affiliation_string":"State Key Laboratory of Com- puter Science , Institute of Software , Chinese Academy of Sciences","institution_ids":["https://openalex.org/I19820366","https://openalex.org/I4210128818","https://openalex.org/I4391767820"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":22,"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":0.7002000212669373,"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":0.7002000212669373,"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":0.19059999287128448,"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/T10028","display_name":"Topic Modeling","score":0.022299999371170998,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.7379999756813049},{"id":"https://openalex.org/keywords/guardrails","display_name":"guardrails","score":0.8790000081062317},{"id":"https://openalex.org/keywords/llm-safety","display_name":"LLM safety","score":0.7820000052452087},{"id":"https://openalex.org/keywords/neuro-symbolic-ai","display_name":"neuro-symbolic AI","score":0.7160000205039978}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.3806562125682831},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.36098039150238037},{"id":"https://openalex.org/C138885662","wikidata":"https://www.wikidata.org/wiki/Q5891","display_name":"Philosophy","level":0,"score":0.11260491609573364}],"mesh":[],"locations_count":3,"locations":[{"id":"pmh:oai:arXiv.org:2402.01822","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.01822","pdf_url":"https://arxiv.org/pdf/2402.01822","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:oai:figshare.com:article/25568364","doi":null,"is_oa":true,"landing_page_url":"https://figshare.com/articles/preprint/Building_guardrails_for_Large_Language_Models/25568364","pdf_url":null,"source":{"id":"https://openalex.org/S4377196282","display_name":"Figshare","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I4210132348","host_organization_name":"Figshare (United Kingdom)","host_organization_lineage":["https://openalex.org/I4210132348"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by-nc-nd","license_id":"https://openalex.org/licenses/cc-by-nc-nd","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Text"},{"id":"doi:10.48550/arxiv.2402.01822","doi":"https://doi.org/10.48550/arxiv.2402.01822","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2402.01822","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2402.01822","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2402.01822","pdf_url":"https://arxiv.org/pdf/2402.01822","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":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G6661291015","display_name":"EnnCore: End-to-End Conceptual Guarding of Neural Architectures","funder_award_id":"EP/T026995/1","funder_id":"https://openalex.org/F4320334627","funder_display_name":"Engineering and Physical Sciences Research Council"}],"funders":[{"id":"https://openalex.org/F4320313467","display_name":"Alan Turing Institute","ror":"https://ror.org/035dkdb55"},{"id":"https://openalex.org/F4320334627","display_name":"Engineering and Physical Sciences Research Council","ror":"https://ror.org/0439y7842"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4391591738.pdf","grobid_xml":"https://content.openalex.org/works/W4391591738.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4409460650","https://openalex.org/W4399912081","https://openalex.org/W4415271252","https://openalex.org/W4411337700","https://openalex.org/W4409350046","https://openalex.org/W7093368538","https://openalex.org/W6948230933","https://openalex.org/W4417353149","https://openalex.org/W7166821409","https://openalex.org/W4416024861"],"abstract_inverted_index":{"As":[0],"Large":[1],"Language":[2],"Models":[3],"(LLMs)":[4],"become":[5],"more":[6,76],"integrated":[7],"into":[8],"our":[9],"daily":[10],"lives,":[11],"it":[12],"is":[13],"crucial":[14],"to":[15,92,119,128,140],"identify":[16],"and":[17,32,67,71,135,138],"mitigate":[18],"their":[19],"risks,":[20],"especially":[21],"when":[22],"the":[23,37,69,72,130,133,142,146],"risks":[24],"can":[25],"have":[26,43],"profound":[27],"impacts":[28],"on":[29,80,98],"human":[30],"users":[31],"societies.":[33],"Guardrails,":[34],"which":[35],"filter":[36],"inputs":[38],"or":[39],"outputs":[40],"of":[41,101,132,145],"LLMs,":[42,96],"emerged":[44],"as":[45],"a":[46,54,89,116],"core":[47],"safeguarding":[48],"technology.":[49],"This":[50],"position":[51],"paper":[52],"takes":[53],"deep":[55],"look":[56],"at":[57],"current":[58],"open-source":[59],"solutions":[60],"(Llama":[61],"Guard,":[62],"Nvidia":[63],"NeMo,":[64],"Guardrails":[65],"AI),":[66],"discusses":[68],"challenges":[70],"road":[73],"towards":[74],"building":[75],"complete":[77],"solutions.":[78],"Drawing":[79],"robust":[81],"evidence":[82],"from":[83],"previous":[84],"research,":[85],"we":[86],"advocate":[87],"for":[88,95],"systematic":[90],"approach":[91],"construct":[93],"guardrails":[94],"based":[97],"comprehensive":[99],"consideration":[100],"diverse":[102],"contexts":[103],"across":[104],"various":[105],"LLMs":[106],"applications.":[107],"We":[108],"propose":[109],"employing":[110],"socio-technical":[111],"methods":[112],"through":[113],"collaboration":[114],"with":[115],"multi-disciplinary":[117],"team":[118],"pinpoint":[120],"precise":[121],"technical":[122],"requirements,":[123,134],"exploring":[124],"advanced":[125],"neural-symbolic":[126],"implementations":[127],"embrace":[129],"complexity":[131],"developing":[136],"verification":[137],"testing":[139],"ensure":[141],"utmost":[143],"quality":[144],"final":[147],"product.":[148]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":13},{"year":2024,"cited_by_count":6}],"updated_date":"2026-10-27T15:51:10.671819","created_date":"2024-02-07T00:00:00"}
