Train a reward model for a coding agent from 100K noisy human scores biased toward short solutions.

Human 1-5 scores are biased toward short, simple code. Debias the labels, fuse test-pass rate without letting the agent game it, and pick PPO with a fused reward over a DPO ranking loss. Plus the hard cap on the action space so the agent can't delete the repo.

Updated Aug 2026 · Grounded in real Forward Deployed Engineer interview loops and written to a senior-engineer editorial bar.