Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.01801 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918366375051264 |
|---|---|
| author | Li, Lehui Wang, Ruining Song, Haochen Mao, Yaoxin Zhang, Tong Wang, Yuyao Fan, Jiayi Zhang, Yitong Ye, Jieping Zhang, Chengqi Gong, Yongshun |
| author_facet | Li, Lehui Wang, Ruining Song, Haochen Mao, Yaoxin Zhang, Tong Wang, Yuyao Fan, Jiayi Zhang, Yitong Ye, Jieping Zhang, Chengqi Gong, Yongshun |
| contents | Automated paper reproduction -- generating executable code from academic papers -- is bottlenecked not by information retrieval but by the tacit knowledge that papers inevitably leave implicit. We formalize this challenge as the progressive recovery of three types of tacit knowledge -- relational, somatic, and collective -- and propose \method, a graph-based agent framework with a dedicated mechanism for each: node-level relation-aware aggregation recovers relational knowledge by analyzing implementation-unit-level reuse and adaptation relationships between the target paper and its citation neighbors; execution-feedback refinement recovers somatic knowledge through iterative debugging driven by runtime signals; and graph-level knowledge induction distills collective knowledge from clusters of papers sharing similar implementations. On an extended ReproduceBench spanning 3 domains, 10 tasks, and 40 recent papers, \method{} achieves an average performance gap of 10.04\% against official implementations, improving over the strongest baseline by 24.68\%. The code will be publicly released upon acceptance; the repository link will be provided in the final version. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_01801 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | What Papers Don't Tell You: Recovering Tacit Knowledge for Automated Paper Reproduction Li, Lehui Wang, Ruining Song, Haochen Mao, Yaoxin Zhang, Tong Wang, Yuyao Fan, Jiayi Zhang, Yitong Ye, Jieping Zhang, Chengqi Gong, Yongshun Artificial Intelligence Automated paper reproduction -- generating executable code from academic papers -- is bottlenecked not by information retrieval but by the tacit knowledge that papers inevitably leave implicit. We formalize this challenge as the progressive recovery of three types of tacit knowledge -- relational, somatic, and collective -- and propose \method, a graph-based agent framework with a dedicated mechanism for each: node-level relation-aware aggregation recovers relational knowledge by analyzing implementation-unit-level reuse and adaptation relationships between the target paper and its citation neighbors; execution-feedback refinement recovers somatic knowledge through iterative debugging driven by runtime signals; and graph-level knowledge induction distills collective knowledge from clusters of papers sharing similar implementations. On an extended ReproduceBench spanning 3 domains, 10 tasks, and 40 recent papers, \method{} achieves an average performance gap of 10.04\% against official implementations, improving over the strongest baseline by 24.68\%. The code will be publicly released upon acceptance; the repository link will be provided in the final version. |
| title | What Papers Don't Tell You: Recovering Tacit Knowledge for Automated Paper Reproduction |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2603.01801 |