metadata
license: apache-2.0
base_model: Qwen/Qwen3.5-9B
tags:
- debugging
- code
- reinforcement-learning
- grpo
- precisecoder
PreciseCoder-9B
馃搫 Paper (arXiv:2610.18755) 路 馃寪 Project page 路 馃捇 Code
Qwen3.5-9B post-trained with reinforcement learning for precise debugging: given buggy Python code, produce a minimal, correct fix (PreciseCoder project).
Recipe. GDPO with two rewards, unit-test pass and unit-gated F1 (edit-block F1 against the ground-truth diff), weighted 2:1.
Eval:
| eval | result |
|---|---|
| PDB test (1205 BigCodeBench + LiveCodeBench debugging tasks) | unit 0.613 路 precision 0.691 路 recall 0.762 路 F1 0.700 路 edit lines 2.9 |
| PDB-Wild (474 tasks: BCB 37 + LCB 209 + SWE-smith 228) | unit 0.323 路 precision 0.429 路 recall 0.474 路 F1 0.432 路 edit lines 12.0 |
| GT Debug (1512 clean inputs) | no-op 50.4% 路 damage 12.9% 路 unit pass 87.1% |
no-op = share of clean (bug-free) inputs returned unchanged; damage = share broken by an edit.
Load with transformers / vLLM (>=0.17) as a standard Qwen3.5 model (thinking enabled by default).
Citation
@article{precisecoder,
title = {{PreciseCoder}: Faithful and Generalizable Debugging Beyond Unit-Test Rewards},
author = {Zhan, Yunkai and Chai, Miaosen and Wang, Shangshang and Zhong, Jike and Yang, Yuqing and Fu, Deqing and Chang, Serina and Zhu, Wang Bill},
journal = {arXiv preprint arXiv:2610.18755},
year = {2026}
}