[Codefuse开源轻训营] Add papers from KDD 2024, 2025

Your task is to find code LLM papers from KDD 2024 and KDD 2025, and add them to the most relevant section following the format of existing entries.

The criteria for inclusion are:
- The paper makes non-trivial contributions to the field of natural language processing, artificial intelligence, or software engineering in one of the following aspects: 1) publicly accessible models, 2) new data resources, 3) novel training/inference algorithms, 4) empirical analysis with novel insights, 5) application to novel tasks.

Use the URL for arXiv abstract page (in the format https://arxiv.org/abs/xxxx.xxxxx) whenever available, since papers from conference sites are not always accessible to everyone.

Review the repository's existing paper entries and section structure, then search KDD 2024 and KDD 2025 for code-related papers meeting the listed contribution criteria. Use arXiv abstract URLs when available and add each qualifying paper to the most relevant section in the existing format; done means the entries are accurate, deduplicated, and consistently formatted.

Written by the indexing model from the issue text.