Text Classification
PEFT
Safetensors
English
Chinese
cybersecurity
cyber-security
security
cyber-defense
blue-team
soc
security-operations
threat-detection
threat-intelligence
cti
alert-triage
incident-response
vulnerability-management
vulnerability-detection
cve
cwe
cvss
mitre-attack
cloud-security
iam
kubernetes-security
devsecops
supply-chain-security
sbom
compliance
ai-security
agent-guardrails
tool-use
function-calling
decision-model
calibration
lora
Eval Results (legacy)
| { | |
| "model": "jue-latest", | |
| "state": { | |
| "headers": { | |
| "From": "\"Microsoft 365 Security\" <no-reply@micros0ft-secure.com>", | |
| "Reply-To": "it-helpdesk@protonmail.example", | |
| "Authentication-Results": "mx.corp.example; spf=fail smtp.mailfrom=micros0ft-secure.com; dkim=none; dmarc=fail header.from=micros0ft-secure.com", | |
| "Subject": "Action required: your mailbox will be suspended in 24 hours" | |
| }, | |
| "body_excerpt": "Click https://micros0ft-secure.com/verify to keep your account active." | |
| }, | |
| "questions": { | |
| "verdict": { | |
| "type": "choice", | |
| "instructions": "Triage verdict for this email.", | |
| "criteria": { | |
| "PHISHING": "Credential phishing", | |
| "SPAM": "Unsolicited marketing", | |
| "BENIGN": "Legitimate email" | |
| } | |
| }, | |
| "dmarc_aligned": { | |
| "type": "noul", | |
| "instructions": "The message passes DMARC alignment." | |
| } | |
| } | |
| } |