https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/Head https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s http://www.nanopub.org/nschema#hasAssertion https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/assertion https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s http://www.nanopub.org/nschema#hasProvenance https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/provenance https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s http://www.nanopub.org/nschema#hasPublicationInfo https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/pubinfo https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://www.nanopub.org/nschema#Nanopublication https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/assertion https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/dc/terms/title KG-DF: A Black-box Defense Framework against Jailbreak Attacks Based on Knowledge Graphs https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/spar/cito/describes https://neverblink.eu/ontologies/llm-kg/methods#KgDf https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/spar/cito/discusses https://neverblink.eu/ontologies/llm-kg/methods#Gcg https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/spar/cito/discusses https://neverblink.eu/ontologies/llm-kg/methods#Pair https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/spar/cito/discusses https://neverblink.eu/ontologies/llm-kg/methods#Ppl https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/spar/cito/discusses https://neverblink.eu/ontologies/llm-kg/methods#Rpo https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/spar/cito/discusses https://neverblink.eu/ontologies/llm-kg/methods#SelfReminder https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/spar/cito/discusses https://neverblink.eu/ontologies/llm-kg/methods#SmoothLlm https://doi.org/10.48550/arXiv.2511.07480 http://purl.org/spar/cito/discusses https://neverblink.eu/ontologies/llm-kg/methods#Tap https://doi.org/10.48550/arXiv.2511.07480 http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://www.w3.org/ns/prov#Entity https://neverblink.eu/ontologies/llm-kg/methods#Gcg http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://purl.org/spar/fabio/Workflow https://neverblink.eu/ontologies/llm-kg/methods#Gcg http://www.w3.org/2000/01/rdf-schema#label GCG https://neverblink.eu/ontologies/llm-kg/methods#KgDf http://purl.org/dc/terms/subject https://neverblink.eu/ontologies/llm-kg/categories#SynergizedReasoning https://neverblink.eu/ontologies/llm-kg/methods#KgDf http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://purl.org/spar/fabio/Workflow https://neverblink.eu/ontologies/llm-kg/methods#KgDf http://www.w3.org/2000/01/rdf-schema#comment KG-DF is a unified framework where a Knowledge Graph (KG) is constructed with safety and general knowledge. An LLM performs semantic parsing of user input to extract keywords, which are then used to retrieve relevant KG triples. These triples are integrated into the LLM's prompt as a "warning" and the LLM then performs a judgment (reasoning) to decide whether to respond or reject, thereby enhancing LLM security against jailbreak attacks and improving general QA. This constitutes a synergistic reasoning process where the LLM acts as an agent interacting with KG-derived knowledge for decision-making. https://neverblink.eu/ontologies/llm-kg/methods#KgDf http://www.w3.org/2000/01/rdf-schema#label KG-DF https://neverblink.eu/ontologies/llm-kg/methods#KgDf https://neverblink.eu/ontologies/llm-kg/hasTopCategory https://neverblink.eu/ontologies/llm-kg/top-categories#SynergizedLLMKG https://neverblink.eu/ontologies/llm-kg/methods#Pair http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://purl.org/spar/fabio/Workflow https://neverblink.eu/ontologies/llm-kg/methods#Pair http://www.w3.org/2000/01/rdf-schema#label PAIR https://neverblink.eu/ontologies/llm-kg/methods#Ppl http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://purl.org/spar/fabio/Workflow https://neverblink.eu/ontologies/llm-kg/methods#Ppl http://www.w3.org/2000/01/rdf-schema#label PPL https://neverblink.eu/ontologies/llm-kg/methods#Rpo http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://purl.org/spar/fabio/Workflow https://neverblink.eu/ontologies/llm-kg/methods#Rpo http://www.w3.org/2000/01/rdf-schema#label RPO https://neverblink.eu/ontologies/llm-kg/methods#SelfReminder http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://purl.org/spar/fabio/Workflow https://neverblink.eu/ontologies/llm-kg/methods#SelfReminder http://www.w3.org/2000/01/rdf-schema#label Self-reminder https://neverblink.eu/ontologies/llm-kg/methods#SmoothLlm http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://purl.org/spar/fabio/Workflow https://neverblink.eu/ontologies/llm-kg/methods#SmoothLlm http://www.w3.org/2000/01/rdf-schema#label SmoothLLM https://neverblink.eu/ontologies/llm-kg/methods#Tap http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://purl.org/spar/fabio/Workflow https://neverblink.eu/ontologies/llm-kg/methods#Tap http://www.w3.org/2000/01/rdf-schema#label TAP https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/provenance https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/assertion http://www.w3.org/ns/prov#wasAttributedTo https://neverblink.eu/ontologies/llm-kg/agent https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/assertion http://www.w3.org/ns/prov#wasDerivedFrom https://doi.org/10.48550/arXiv.2511.07480 https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/pubinfo https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s http://purl.org/dc/terms/created 2026-02-26T15:46:51.325Z https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s http://purl.org/dc/terms/creator https://neverblink.eu/ontologies/llm-kg/agent https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s http://purl.org/nanopub/x/hasNanopubType https://neverblink.eu/ontologies/llm-kg/PaperAssessmentResult https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s http://www.w3.org/2000/01/rdf-schema#label LLM-KG assessment for paper 10.48550/arXiv.2511.07480 https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/sig http://purl.org/nanopub/x/hasAlgorithm RSA https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/sig http://purl.org/nanopub/x/hasPublicKey MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAwNz2QK3SEifno78S7+48zUB0xpTex3mAzW73ZimHqNcdEMU5/apslrGrTHGFAt/Chocgo++r6JQp5ygY7NyJHGWdaIqnt85pjX4PbNfLAvapyUO00qZP34fY61w4eZ9UMtleWEsmZKRtQPyJ8ODl46i/rfPuZlcJGpM9Nmy5mpGWuepqIEvF4a/t7pLVeCEDFSYXT+yaiygt6ynIK5f7TtEDhZpeUf/Q74WhMPJXm4yTU/hqOX4IW+50kWHNArGGZwUaXwzyG6M3Zd6UMModryGkLqS4H/MSE3ZA1Ylnms7BfWLEXhMWlaKi6HRV4nGRDLhxVSi9LSRi3LWKLhNIIQIDAQAB https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/sig http://purl.org/nanopub/x/hasSignature lbUpkFyvtlToOFTDNQzm0tUOaYlqOfLMayIC7OhYZlCkq5mbsCQnFQ69m6FPO9YtAzU/QV/+LUWC2cJE17eMXWGrMiJHVElFwcVmdab6MxyrE+Sz8q9R5zGs0G55+ZL4zP+rWJsBkQSdXxUl07mx/Fvz4RvGalKZ9j+n1wmqq3NDUWpMphjvbPqn/ysH793uUNDtyIg5DUr2XCcM1iSgmfndZ05BGeQM0YqWxDDo1Nn19RHftcs3TI51qj4BBfOQWECu2ucT0QGZc+dWoQvvl00p9gOb2cDK+vrds8oMQSw3fqkbi1y81pKJ18CQNwmimWru0IGZSoHVPjtlvAL2mQ== https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/sig http://purl.org/nanopub/x/hasSignatureTarget https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s https://w3id.org/np/RAe5-l8XUIWC0J0kshCll4i11gVSeDXeeCpMnD_bZZ13s/sig http://purl.org/nanopub/x/signedBy https://neverblink.eu/ontologies/llm-kg/agent