RepoForge: Training a SOTA Fast-thinking SWE Agent with an End-to-End Data Curation Pipeline Synergizing SFT and RL at Scale. Chen, Z., Zhao, C., Chen, B., Lin, D., Chen, Y., Leung, A., Rajbahadur, G. K., Oliva, G. A., Zhang, H., Bhatia, A., Yong, C. C., & Hassan, A. E. CoRR, 2025.
Paper doi bibtex @article{DBLP:journals/corr/abs-2508-01550,
author = {Zhilong Chen and
Chengzong Zhao and
Boyuan Chen and
Dayi Lin and
Yihao Chen and
Arthur Leung and
Gopi Krishnan Rajbahadur and
Gustavo Ansaldi Oliva and
Haoxiang Zhang and
Aaditya Bhatia and
Chong Chun Yong and
Ahmed E. Hassan},
title = {RepoForge: Training a {SOTA} Fast-thinking {SWE} Agent with an End-to-End
Data Curation Pipeline Synergizing {SFT} and {RL} at Scale},
journal = {CoRR},
volume = {abs/2508.01550},
year = {2025},
url = {https://doi.org/10.48550/arXiv.2508.01550},
doi = {10.48550/ARXIV.2508.01550},
eprinttype = {arXiv},
eprint = {2508.01550},
timestamp = {Wed, 10 Sep 2025 01:00:00 +0200},
biburl = {https://dblp.org/rec/journals/corr/abs-2508-01550.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}
Downloads: 0
{"_id":"ANWjaAXDzdXgxe55f","bibbaseid":"chen-zhao-chen-lin-chen-leung-rajbahadur-oliva-etal-repoforgetrainingasotafastthinkingsweagentwithanendtoenddatacurationpipelinesynergizingsftandrlatscale-2025","author_short":["Chen, Z.","Zhao, C.","Chen, B.","Lin, D.","Chen, Y.","Leung, A.","Rajbahadur, G. K.","Oliva, G. A.","Zhang, H.","Bhatia, A.","Yong, C. C.","Hassan, A. E."],"bibdata":{"bibtype":"article","type":"article","author":[{"firstnames":["Zhilong"],"propositions":[],"lastnames":["Chen"],"suffixes":[]},{"firstnames":["Chengzong"],"propositions":[],"lastnames":["Zhao"],"suffixes":[]},{"firstnames":["Boyuan"],"propositions":[],"lastnames":["Chen"],"suffixes":[]},{"firstnames":["Dayi"],"propositions":[],"lastnames":["Lin"],"suffixes":[]},{"firstnames":["Yihao"],"propositions":[],"lastnames":["Chen"],"suffixes":[]},{"firstnames":["Arthur"],"propositions":[],"lastnames":["Leung"],"suffixes":[]},{"firstnames":["Gopi","Krishnan"],"propositions":[],"lastnames":["Rajbahadur"],"suffixes":[]},{"firstnames":["Gustavo","Ansaldi"],"propositions":[],"lastnames":["Oliva"],"suffixes":[]},{"firstnames":["Haoxiang"],"propositions":[],"lastnames":["Zhang"],"suffixes":[]},{"firstnames":["Aaditya"],"propositions":[],"lastnames":["Bhatia"],"suffixes":[]},{"firstnames":["Chong","Chun"],"propositions":[],"lastnames":["Yong"],"suffixes":[]},{"firstnames":["Ahmed","E."],"propositions":[],"lastnames":["Hassan"],"suffixes":[]}],"title":"RepoForge: Training a SOTA Fast-thinking SWE Agent with an End-to-End Data Curation Pipeline Synergizing SFT and RL at Scale","journal":"CoRR","volume":"abs/2508.01550","year":"2025","url":"https://doi.org/10.48550/arXiv.2508.01550","doi":"10.48550/ARXIV.2508.01550","eprinttype":"arXiv","eprint":"2508.01550","timestamp":"Wed, 10 Sep 2025 01:00:00 +0200","biburl":"https://dblp.org/rec/journals/corr/abs-2508-01550.bib","bibsource":"dblp computer science bibliography, https://dblp.org","bibtex":"@article{DBLP:journals/corr/abs-2508-01550,\n author = {Zhilong Chen and\n Chengzong Zhao and\n Boyuan Chen and\n Dayi Lin and\n Yihao Chen and\n Arthur Leung and\n Gopi Krishnan Rajbahadur and\n Gustavo Ansaldi Oliva and\n Haoxiang Zhang and\n Aaditya Bhatia and\n Chong Chun Yong and\n Ahmed E. Hassan},\n title = {RepoForge: Training a {SOTA} Fast-thinking {SWE} Agent with an End-to-End\n Data Curation Pipeline Synergizing {SFT} and {RL} at Scale},\n journal = {CoRR},\n volume = {abs/2508.01550},\n year = {2025},\n url = {https://doi.org/10.48550/arXiv.2508.01550},\n doi = {10.48550/ARXIV.2508.01550},\n eprinttype = {arXiv},\n eprint = {2508.01550},\n timestamp = {Wed, 10 Sep 2025 01:00:00 +0200},\n biburl = {https://dblp.org/rec/journals/corr/abs-2508-01550.bib},\n bibsource = {dblp computer science bibliography, https://dblp.org}\n}\n\n","author_short":["Chen, Z.","Zhao, C.","Chen, B.","Lin, D.","Chen, Y.","Leung, A.","Rajbahadur, G. K.","Oliva, G. A.","Zhang, H.","Bhatia, A.","Yong, C. C.","Hassan, A. E."],"key":"DBLP:journals/corr/abs-2508-01550","id":"DBLP:journals/corr/abs-2508-01550","bibbaseid":"chen-zhao-chen-lin-chen-leung-rajbahadur-oliva-etal-repoforgetrainingasotafastthinkingsweagentwithanendtoenddatacurationpipelinesynergizingsftandrlatscale-2025","role":"author","urls":{"Paper":"https://doi.org/10.48550/arXiv.2508.01550"},"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://dblp.org/pid/187/9420.bib","dataSources":["FLwtLJyrjR3ZC72vq"],"keywords":[],"search_terms":["repoforge","training","sota","fast","thinking","swe","agent","end","end","data","curation","pipeline","synergizing","sft","scale","chen","zhao","chen","lin","chen","leung","rajbahadur","oliva","zhang","bhatia","yong","hassan"],"title":"RepoForge: Training a SOTA Fast-thinking SWE Agent with an End-to-End Data Curation Pipeline Synergizing SFT and RL at Scale","year":2025}