{"acceptedBy":{"actor":"ai","name":null,"on":"2026-10-22"},"aliases":[],"change":{"last4":1,"percent":null,"prior4":0},"description":{"editedBy":null,"state":"none"},"href":"/wiki/model-yufeng-xguard-reason-8b/","itemCount":1,"key":"model--yufeng-xguard-reason-8b","kind":"model","name":"YuFeng-XGuard-Reason-8B","neighbours":[{"key":"dataset--toxic-chat","kind":"dataset","name":"toxic-chat","relation":"mentioned_with","weight":1},{"key":"licence--apache-2-0","kind":"licence","name":"Apache-2.0","relation":"mentioned_with","weight":1},{"key":"licence--cc-by-nc-4-0","kind":"licence","name":"CC-BY-NC-4.0","relation":"mentioned_with","weight":1},{"key":"maker--accuknox-inc","kind":"maker","name":"AccuKnox Inc.","relation":"mentioned_with","weight":1},{"key":"model--qwen3-5-4b","kind":"model","name":"Qwen3.5-4B","relation":"mentioned_with","weight":1},{"key":"platform--hugging-face","kind":"platform","name":"Hugging Face","relation":"mentioned_with","weight":1},{"key":"task--content-moderation","kind":"task","name":"content moderation","relation":"mentioned_with","weight":1},{"key":"task--prompt-injection-defence","kind":"task","name":"prompt injection defence","relation":"mentioned_with","weight":1},{"key":"technique--lora","kind":"technique","name":"LoRA","relation":"mentioned_with","weight":1}],"relatedItems":[{"publishedAt":"2026-10-16T00:00:00Z","reason":"model--yufeng-xguard-reason-8b","title":"AccuKnox Inc. releases Guard 1.1, an Apache-2.0 guardrail for agent inputs and prompt injection","url":"/decisions/accuknox-guard-1-1/"}],"schemaVersion":1,"series26":[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0],"series8":[0,0,0,0,0,0,1,0],"weeksOfHistory":2}