Deception Benchmark

Named by 2 kept items in this workspace.

Search every workspace for Deception Benchmark. It matches the name as written, so it can find other things that share it.

The items that name it 2

AWS Continuum sets a new standard in autonomous code security
article2026-10-05
AWS evaluation that tests how reliably individual frontier models distinguish real vulnerabilities from safe code, complementing the system-level CyberGym-E2E evaluation
The state of AI for security: Measuring what matters most for building trust
article2026-09-09
new AWS-released benchmark (14,822 samples, 16 languages, 70+ CWE categories) measuring whether AI models can distinguish real vulnerabilities from safe code that looks risky but has mitigations closing the exploit path; includes code-level (6,988 samples) and environment-gated (2,707 samples) chall

All technologies in this workspace.