seccodeplt-qwen2.5-coder-7b-grpo-kl-beta-0.001-real-reward-v2
GRPO with KL regularization (beta=0.001) for the SecCodePLT+ compliance experiment using
Qwen/Qwen2.5-Coder-7B-Instruct. This v2 run corrects causal-label alignment and uses the
official ReaL safety-unit-test reward with DAPO-style token loss and dynamic
sampling. Training used seed 42 and the official 655-example training split.
Evaluation used greedy decoding on all 164 official test examples.
Evaluation
| Metric | Value |
|---|---|
| Mean reward | 0.511317 |
| Output format pass | 98.78% |
| Syntax pass | 98.17% |
| Capability pass | 38.41% |
| Safety pass | 64.02% |
| Joint pass | 31.10% |
Limitations
This is a single-seed research checkpoint evaluated with the benchmark's resource-bounded Python verifier. It is not a general guarantee of secure code.