Could CVE rates set a security bar for code-generating models?
_arohan_ · x · 2026-11-07
Anthropic researcher arohan proposes tracing which models generated code that later produced CVEs, and using vulnerability rates to set a minimum security bar for models used in foundational software. Without such a mechanism, he argues, market pressure naturally optimizes for cheaper tokens rather than safer models.
Related event: Researchers Propose CVE Rate as Safety Threshold for Code Models(3 posts)→
More from Models
- ChatGPT reportedly mass-banning accounts, long-time paying users included — zyyyyl9144 · 2026-11-07
- Dev runs 284B-parameter DeepSeek-V4-Flash on CPU only, squeezed into 64GB DDR5 RAM — solyarisoftware · 2026-11-07
- Inkling 2 Small jumps +240 pts in Code Arena: WebDev, would be #1 US open model — SonglinYang4 · 2026-11-07
- MIT's Vector Policy Optimization replaces scalar RL rewards to boost LLM test-time search, accepted at NeurIPS 2026 — kastnerkyle · 2026-11-07
- Uncensored DeepSeek-distilled Qwen3.5-9B merge hits HF trending — xtools-at · 2026-11-07
- Anthropic paper: 3M tokens of moral dilemmas cut model malicious behavior to 0%, 28x efficiency — garrytan · 2026-11-07