Yiqing Xie
Language Technologies Institute, CMU. yiqingxi@andrew.cmu.edu
Rm 6413, Gates and Hillman Centers
4902 Forbes Ave
Pittsburgh, PA 15213, USA
I am a final year Ph.D. student at the Language Technologies Institute of Carnegie Mellon University and I am advised by Carolyn Rosé and Daniel Fried on agentic post-training and coding agents. Previously, I obtained my Master degree in the data mining group at the University of Illinois Urbana-Champaign supervised by Jiawei Han and obtained my Bachelor degree in Hong Kong University of Science and Technology, where I received the Academic Achievement Medal.
My research mainly focuses on agentic post-training, coding agents, and factuality and reliability:
-
Agentic Post-Training. I construct scalable and effective agent training environments, design lightweight training tasks that generalize to multiple coding agent benchmarks, and tackle reward-hacking for rubric-based RL.
Hybrid-Gym (ICML’26) · RepoST (COLM’25) · USI (COLM’26) · Ditto (preprint) · FORK (preprint) -
Coding Agents. I develop multi-agent inference settings that are effective, flexible, and cost-efficient, design automatic analysis methods that efficiently discover the influential factors of coding agent training demonstrations, and build evaluation benchmarks for realistic workplace tasks.
Steer (EMNLP’26 Oral) · CATRadar (preprint) · Hybrid-Gym (ICML’26) · RepoST (COLM’25) · TAC (Neurips’25) · CodeRAG-Bench (NAACL Findings’25) -
Factuality and Reliability. I train critique-based evaluators that substantially improve factuality training, design evaluation frameworks for reliable and trustworthy clinical text generation, and continuous pretrain LLMs for better generalizability.
FenCE (ACL’25) · DocLens (ACL’24) · AnchorDR (SIGIR’23) · METRO-T0 (ACL’23)
👩💻 I am on the industry job market in 2027! Please check out my CV.
News
| Oct 14, 2026 | New preprints on coding agent training effectiveness analysis (CATRadar) and mitigating reward hacking in Rubric-based RL (FORK)!!! |
|---|---|
| Sep 4, 2026 | “Steer, Don’t Solve” got accepted to EMNLP 2026!! (Steer) |
| Jul 8, 2026 | Two papers got accepted to COLM 2026!! Congrats to my wonderful collaborators! (MetaLint and USI) |
| Jun 5, 2026 | Gave a talk about designing scalable synthetic training tasks for coding agents at Nvidia! |
| May 25, 2026 | Some of my recent preprints: Training small critic models to steer large coding agents (Steer); Training automatic linting models (MetaLint); Analysis on LLM-based user simulation (USI); Human behavior simulation training (DITTO). |
| May 15, 2026 | Proposed my PhD thesis on “Synthetic Training Data Construction for Coding Agents“🦸🏻♀️!!! |
| Apr 30, 2026 | Hybrid-Gym got accepted to ICML 2026! (Hybrid-Gym) |
| Nov 11, 2025 | Gave a guest lecture in CMU 11-891 Neural code generation! (slides) |
| Aug 26, 2025 | Start TA-ing for 11-891 Neural code generation!! |
| Aug 20, 2025 | Two papers on repo-level code generation got accepted to EMNLP 2025!! (SACL, Strong-Weak-Colab) |
Educations
Work experience
- Meta AI
2024.05 - 2024.10 - Research Intern, GenAI
- Work on training a fine-grained critic-based evaluator model and use it to improve generators' factuality [FenCE, ACL'25]
- Manager: Hejia Zhang; Peers: Di Jin, Sinong Wang
- Microsoft Research Redmond
2023.06 - 2023.08 - Research Intern, Health Futures
- Work on a multi-aspect fine-grained evaluation framework of medical text generation [DocLens, ACL'24]
- Manager: Sheng Zhang, Hao Cheng, Hoifung Poon
- Microsoft Research Redmond
2022.05 - 2022.08 - Research Intern, Productivity and Intelligence group
- Work on continuously pre-trained models for zero-shot dense retrieval [Anchor-DR, SIGIR'23]
- Manager: Chenyan Xiong
- Alibaba DAMO Academy
2020.07 - 2021.02 - Research Intern, Data Analytics and Intelligence Lab
- Work on few-shot interaction recommendation under multiple scenarios [KoMen, WWW'22]
- Manager: Yaliang Li, Bolin Ding
Invited Talks
- Designing Scalable and Generalizable Synthetic Training Tasks for Coding Agents
June 2026 - Nvidia Reading Club, Nvidia
- Synthetic Data Approaches for Coding Agent Training
Nov 2025 - Guest Lecture for 11-891, Neural code generation, Carnegie Mellon University
- Constructing Scalable Repository-Level Environments for Coding Agent Training
Apr 2025 - EFML Reading Group, Stanford / University of Washington
- Multi-aspect Fine-grained Evaluation for Medical Text Generation
Apr 2024 - Health and Life Science Seminar, Microsoft Research
Student Mentoring
- Shubham Gandhi
Sept. 2024 - Present - CMU MIIS Master's Student → CMU Ph.D. Student
- Dhruv Gupta
Dec. 2024 - May 2025 - CMU MIIS Master's Student → Skild AI
- Gayathri Ganesh Lakshmy
Dec. 2024 - May 2025 - CMU MIIS Master's Student → Databricks
- Gaokai Zhang
Nov. 2025 - Present - CMU MIIS Master's Student
- Ruichen Zhu
May 2026 - Aug. 2026 - CMU Undergraduate Student
- Hyun Ryu
Aug. 2026 - Oct. 2026 - KAIST Master's Student
- Jerry Yan
Sept. 2026 - Present - CMU MCDS Master's Student
Honors and Awards
Additional Information
Selected publications
- preprintSuccessful Rollouts are Not Equal: Discovering Influential Behaviors for Training Coding Agents(preprint, 2026)
- EMNLPSteer, Don’t Solve: Training Small Critic Models for Large Code AgentsIn Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP, 2026)
- ICMLHybrid-Gym: Training Coding Agents to Generalize Across TasksIn Forty-third International Conference on Machine Learning (ICML, 2026)
- COLMRepoST: Scalable Repository-Level Coding Environment Construction with Sandbox TestingIn Proceedings of the 2nd Conference on Language Modeling (COLM, 2025)
- ACLImproving Model Factuality with Fine-grained Critique-based EvaluatorIn Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL, 2025)
- ACLDocLens: Multi-aspect Fine-grained Evaluation for Medical Text GenerationIn Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (ACL, 2024)
- EMNLP FindingsData Augmentation for Code Translation with Comparable Corpora and Multiple ReferencesIn Findings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP Findings, 2023)
- SIGIRUnsupervised Dense Retrieval Training with Web AnchorsIn Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR, 2023)
- ACL FindingsEider: Evidence-enhanced Document-level Relation ExtractionIn Findings of the 60th Annual Meeting of the Association for Computational Linguistics (ACL Findings, 2022)
- WWWKoMen: Domain Knowledge-Guided Few-Shot Interaction Recommendation on Multiplex NetworksIn Proceedings of the Web Conference (WWW, 2022)
- IJCAIWhen Do GNNs Work: Understanding and Improving Neighborhood AggregationIn Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence (IJCAI, 2020)
- WWWGuiding Corpus-Based Set Expansion by Auxiliary Sets Generation and Co-ExpansionIn Proceedings of The Web Conference (WWW, 2020)