wuyoscar/ISC-Bench

[NeurIPS 2026] Safety failures in coding agents and frontier LLMs. Reproducible workspaces for studying when AI agents generate unsafe data during task completion.

4 best points · 1 comments · 1 Hacker News post · 2,606 GitHub stars · View on GitHub

Hacker News discussions

• A Big Alignment Loophole of Current Froniter LLMs (4 points, 1 comments, by pythonsen, 2026-04-04)