GLM-5.3 finds critical vulnerability in Cursor; post-training boosts scores
rohanpaul_ai · x · 2026-08-16
During a complex reverse-engineering task on Cursor, GLM-5.3 identified a "potentially serious vulnerability," which has been privately disclosed with a fix in progress. Security benchmarks show significant gains: CyberGym score rose to 84.5%, and ExploitBench more than doubled from 24.4% to 54.4%. Z.ai noted that all performance improvements came from scaling post-training, avoiding costly pretraining runs.
Related event: GLM-5.3 Reverse-Engineering Uncovers Serious Cursor Vulnerability(3 posts)→
More from Models
- TokenSpeed MLA 2 hits Vera Rubin: up to 3.44× faster MLA decode for Kimi K3 — zhyncs42 · 2026-10-16
- 200 chess games suggest in-context learning doesn't work yet, experimenter says — akbirthko · 2026-10-16
- OpenAI's internal model shows real math "research taste," experts say its proofs go beyond prior literature — cong_ml · 2026-10-16
- User catches Opus spinning up throwaway zz- debug scripts, echoing the METR OAI-HuggingFace incident — holdenmatt · 2026-10-16
- OpenAI drops 722 math papers tied to unreleased model, mathematicians demand receipts — thursdai_pod · 2026-10-16
- Fable 5/5.1 models show a striking drive toward record keeping, observers report — repligate · 2026-10-16