@inproceedings{gandhi2026codeslop,title={Evaluating Code Slop in Long-Horizon Coding Agents},author={Gandhi, Shubham and Agarwal, Shyam and Kotalwar, Nachiket and Naik, Atharva},booktitle={Workshop on Agent Behavior (WAB) at the Conference on Language Modeling (COLM)},year={2026},month=oct,}
@inproceedings{gandhi2026steer,title={Steer, Don't Solve: Training Small Critic Models for Large Code Agents},author={Gandhi, Shubham and Xie, Yiqing and Naik, Atharva and Zhu, Ruichen and Ros{\'e}, Carolyn},booktitle={Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP)},year={2026},month=nov,}
@inproceedings{xie2026hybridgym,title={Hybrid-Gym: Training Coding Agents to Generalize Across Tasks},author={Xie, Yiqing and Liu, Emmy and Zhang, Gaokai and Kotalwar, Nachiket and Gandhi, Shubham and Acharya, Sathwik and Wang, Xingyao and Ros{\'e}, Carolyn and Neubig, Graham and Fried, Daniel},booktitle={Proceedings of the 43rd International Conference on Machine Learning (ICML)},year={2026},month=jul,}
@inproceedings{gandhi2025astray,title={When Agents go Astray: Course-Correcting SWE Agents with PRMs},author={Gandhi, Shubham and Tsay, Jason and Ganhotra, Jatin and Kate, Kiran and Rizk, Yara},booktitle={Scaling Environments for Agents (SEA) Workshop at NeurIPS},year={2025},month=dec,}
@techreport{naik2025secure,title={Secure and Useful Models are Reasonable: Aligning Code Models via Utility-Preserving Reasoning},author={Naik, Atharva and Xie, Alex and Rao, Abhinav and Agarwal, Anmol and Gandhi, Shubham and Hilton, Michael and Ros{\'e}, Carolyn},institution={Amazon Nova AI Challenge Proceedings (Trusted AI Track, Finalist)},year={2025},month=aug,}
@inproceedings{gandhi2025strongweak,title={An Empirical Study on Strong-Weak Model Collaboration for Repo-level Code Generation},author={Gandhi, Shubham and Naik, Atharva and Xie, Yiqing and Ros{\'e}, Carolyn},booktitle={Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP)},year={2025},month=nov,}
@inproceedings{gandhi2025researchcodeagent,title={ResearchCodeAgent: An LLM Multi-Agent System for Automated Codification of Research Methodologies},author={Gandhi, Shubham and Shah, Dhruv and Patwardhan, Manasi and Vig, Lovekesh and Shroff, Gautam},booktitle={AI4Research Workshop at AAAI},year={2025},month=mar,}
@inproceedings{gandhi2024budgetmlagent,title={BudgetMLAgent: A Cost-Effective LLM Multi-Agent System for Automating Machine Learning Tasks},author={Gandhi, Shubham and Patwardhan, Manasi and Vig, Lovekesh and Shroff, Gautam},booktitle={Proceedings of the 4th International Conference on AI-ML Systems (AIMLSystems)},year={2024},month=oct,doi={10.1145/3703412.3703416},}
@inproceedings{gandhi2024cobol,title={Translation of Low-Resource COBOL to Logically Correct and Readable Java leveraging High-Resource Java Refinement},author={Gandhi, Shubham and Patwardhan, Manasi and Khatri, Jyotsana and Vig, Lovekesh and Medicherla, Raveendra Kumar},booktitle={LLM4Code Workshop at the International Conference on Software Engineering (ICSE)},year={2024},month=apr,doi={10.1145/3643795.3648388},}
Preprints
2026
Preprint
Ulysses: A Security Harness for Synthesizing Verifiable Web Exploits
Aditya Oke, André Vicente Duarte, Rui Melo, and 6 more authors
@article{oke2026ulysses,title={Ulysses: A Security Harness for Synthesizing Verifiable Web Exploits},author={Oke, Aditya and Duarte, Andr{\'e} Vicente and Melo, Rui and Gandhi, Shubham and Kotalwar, Nachiket and Khandor, Charmi and Wang, Danqing and Ros{\'e}, Carolyn and Li, Lei},journal={arXiv preprint arXiv:2610.31831},year={2026},month=oct,}
Preprint
Self-Evolving Static Analysis via LLM Pseudo-Labeling
Rui Melo, André Vicente Duarte, Aditya Oke, and 5 more authors
@article{melo2026selfevolving,title={Self-Evolving Static Analysis via LLM Pseudo-Labeling},author={Melo, Rui and Duarte, Andr{\'e} Vicente and Oke, Aditya and Gandhi, Shubham and Kotalwar, Nachiket and Abreu, Rui and Li, Lei and Ros{\'e}, Carolyn},journal={arXiv preprint arXiv:2610.30184},year={2026},month=oct,}
Preprint
Exploring Visual Access and Proactive AI Tutoring for Handwritten Mathematics
Nachiket Kotalwar, Conrad Borchers, Shubham Gandhi, and 3 more authors
@article{kotalwar2026tutoring,title={Exploring Visual Access and Proactive AI Tutoring for Handwritten Mathematics},author={Kotalwar, Nachiket and Borchers, Conrad and Gandhi, Shubham and Purkayastha, Partha Sarathi and Yaron, David and Ros{\'e}, Carolyn P.},journal={arXiv preprint arXiv:2610.26411},year={2026},month=oct,}
Preprint
ABSENTIA: Detecting Broken Access Control Vulnerabilities in Web Applications
André Vicente Duarte, Aditya Oke, Rui Melo, and 7 more authors
arXiv preprint arXiv:2610.00977 (under review), Oct 2026
@article{duarte2026absentia,title={ABSENTIA: Detecting Broken Access Control Vulnerabilities in Web Applications},author={Duarte, Andr{\'e} Vicente and Oke, Aditya and Melo, Rui and Gandhi, Shubham and Kotalwar, Nachiket and Khandor, Charmi and Wang, Danqing and Oliveira, Arlindo L. and Ros{\'e}, Carolyn and Li, Lei},journal={arXiv preprint arXiv:2610.00977 (under review)},year={2026},month=oct,}
Preprint
DRACO: Fine-Grained Credit Assignment with Dynamic Rubrics for Long-Horizon Agent Training
Shubham Gandhi*, Saurabh Goyal*, Kiran Kate, and 1 more author
@article{gandhi2026draco,title={DRACO: Fine-Grained Credit Assignment with Dynamic Rubrics for Long-Horizon Agent Training},author={Gandhi, Shubham and Goyal, Saurabh and Kate, Kiran and Rizk, Yara},journal={arXiv preprint arXiv:2609.04094 (under review)},year={2026},month=sep,}
2023
Preprint
StaticFixer: From Static Analysis to Static Repair
Naman Jain, Shubham Gandhi, Atharv Sonwane, and 5 more authors
@article{jain2023staticfixer,title={StaticFixer: From Static Analysis to Static Repair},author={Jain, Naman and Gandhi, Shubham and Sonwane, Atharv and Kanade, Aditya and Natarajan, Nagarajan and Parthasarathy, Suresh and Rajamani, Sriram and Sharma, Rahul},journal={arXiv preprint arXiv:2307.12465},year={2023},month=jul,}