Developer's Hand-Written Code Becomes Third-Hardest Task in FrontierCS 2 Benchmark

RAGEN, code hand-written by developer wzenus in early 2025 before coding agents became popular, ranks as the third-hardest task in FrontierCS 2, requiring a 0.5B model to improve RL performance within just 0.5 A100 GPU-hours.

2026-10-30 ~ 2026-10-30 · 4 related posts

• RAGEN ranks 3rd hardest in FrontierCS-2 — built from code hand-written in Jan 2025 — wzenus · 2026-10-30
• Why RAGEN is so hard: lifting a 0.5B model's RL perf in 0.5 A100 GPU hours — wzenus · 2026-10-30
• RAGEN ranks 3rd hardest in FrontierCS 2, built on hand-written code — wzenus · 2026-10-30
• RAGEN ranks 3rd hardest in FrontierCS 2: lift a 0.5B model with 0.5 A100 hour — wzenus · 2026-10-30