The authors first analyzed memory-augmented repository repair and identified three problems: repair histories are unevenly distributed across repositories, larger memory stores do not consistently improve success, and stored experiences are often concentrated in some repair stages but scarce in others.
Program repair improves when memories match repository needs and repair stages
AdaRepair-Mem selects historical repair experiences by coverage, quality, and workflow stage rather than relying on same-repository retrieval alone.
Chinese Tech
Z. C. Luo · J. C. Guo · W. J. He · S. Y. Wang · J. C. Yu · F. M. Zhao · +7 more
Zhejiang University · Tencent
Research Digest··2 min read
Thread:Memory Management for Agents
Luo and colleagues developed an adaptive memory-retrieval framework for large-language-model-based repository repair.
Why this paper
From Tencent and Zhejiang University · Part of Memory Management for Agents, now 37 papers
In one line
Adaptive retrieval of repair experiences by coverage, quality, and stage improves repository-level program repair.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ·No benchmark numbers found
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§