publications

2026

1. EMNLP

   Randomized YaRN Improves Length Generalization for Long-Context Reasoning

   Manas Mehta, Fangcong Yin, and Greg Durrett

   In Findings of the Association for Computational Linguistics: EMNLP 2026, 2026

   arXiv Bib

   @inproceedings{mehta2026yarn,
     title = {Randomized YaRN Improves Length Generalization for Long-Context Reasoning},
     author = {Mehta, Manas and Yin, Fangcong and Durrett, Greg},
     booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2026},
     year = {2026},
   }

2. COLM

   Detecting and Suppressing Reward Hacking with Gradient Fingerprints

   Songtao Wang, Quang Hieu Pham, Fangcong Yin, Xinpeng Wang, Jocelyn Qiaochu Chen, Greg Durrett, and Xi Ye

   In Proceedings of the Conference on Language Modeling (COLM), 2026

   arXiv Bib

   @inproceedings{wang2026reward,
     title = {Detecting and Suppressing Reward Hacking with Gradient Fingerprints},
     author = {Wang, Songtao and Pham, Quang Hieu and Yin, Fangcong and Wang, Xinpeng and Chen, Jocelyn Qiaochu and Durrett, Greg and Ye, Xi},
     booktitle = {Proceedings of the Conference on Language Modeling (COLM)},
     year = {2026},
   }

3. arXiv

   Visually Grounded Self-Reflection for Vision-Language Models via Reinforcement Learning

   Liyan Tang^*, Fangcong Yin^*, and Greg Durrett

   Preprint, 2026

   arXiv Bib

   @article{tang2026reflection,
     title = {Visually Grounded Self-Reflection for Vision-Language Models via Reinforcement Learning},
     author = {Tang, Liyan and Yin, Fangcong and Durrett, Greg},
     journal = {Preprint},
     year = {2026},
   }

4. arXiv

   DySCO: Dynamic Attention-Scaling Decoding for Long-Context LMs

   Xi Ye^*, Wuwei Zhang^*, Fangcong Yin, Howard Yen, and Danqi Chen

   arXiv preprint arXiv:2602.22175, 2026

   arXiv Bib

   @article{ye2026dysco,
     title = {DySCO: Dynamic Attention-Scaling Decoding for Long-Context LMs},
     author = {Ye, Xi and Zhang, Wuwei and Yin, Fangcong and Yen, Howard and Chen, Danqi},
     journal = {arXiv preprint arXiv:2602.22175},
     year = {2026},
   }

5. arXiv

   When Many Answers Are Valid, Voting Fails: Symbolic Verification for Best-of-K Causal Reasoning in LLMs

   Omatharv Bharat Vaidya, Connor Thomas Jerzak, Zayne Sprague, Fangcong Yin, and Nhat Ho

   arXiv preprint arXiv:2608.03506, 2026

   arXiv Bib

   @article{vaidya2026voting,
     title = {When Many Answers Are Valid, Voting Fails: Symbolic Verification for Best-of-K Causal Reasoning in LLMs},
     author = {Vaidya, Omatharv Bharat and Jerzak, Connor Thomas and Sprague, Zayne and Yin, Fangcong and Ho, Nhat},
     journal = {arXiv preprint arXiv:2608.03506},
     year = {2026},
   }

2025

1. EMNLP

   Learning Composable Chains-of-Thought

   Fangcong Yin, Zeyu Leo Liu, Liu Leqi, Xi Ye, and Greg Durrett

   In Findings of the Association for Computational Linguistics: EMNLP 2026. Also presented at the Workshop on Foundations of Reasoning in Language Models (Oral), NeurIPS 2025 , 2025

   arXiv Bib

   @inproceedings{yin2025composable,
     title = {Learning Composable Chains-of-Thought},
     author = {Yin, Fangcong and Liu, Zeyu Leo and Leqi, Liu and Ye, Xi and Durrett, Greg},
     booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2026},
     year = {2025},
   }

2. EMNLP

   Query-Focused Retrieval Heads Improve Long-Context Reasoning and Re-ranking

   Wuwei Zhang, Fangcong Yin, Howard Yen, Danqi Chen, and Xi Ye

   In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP), 2025

   arXiv Bib

   @inproceedings{zhang2025qrhead,
     title = {Query-Focused Retrieval Heads Improve Long-Context Reasoning and Re-ranking},
     author = {Zhang, Wuwei and Yin, Fangcong and Yen, Howard and Chen, Danqi and Ye, Xi},
     booktitle = {Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP)},
     year = {2025},
   }

3. NeurIPS

   ChartMuseum: Testing Visual Reasoning Capabilities of Large Vision-Language Models

   Liyan Tang, Grace Kim, Xinyu Zhao, Thom Lake, Wenxuan Ding, Fangcong Yin, Prasann Singhal, Manya Wadhwa, Zeyu Leo Liu, Zayne Sprague, Ramya Namuduri, Bodun Hu, Juan Diego Rodriguez, Puyuan Peng, and Greg Durrett

   In Proceedings of the 39th Conference on Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track, 2025

   arXiv Bib

   @inproceedings{tang2025chartmuseum,
     title = {ChartMuseum: Testing Visual Reasoning Capabilities of Large Vision-Language Models},
     author = {Tang, Liyan and Kim, Grace and Zhao, Xinyu and Lake, Thom and Ding, Wenxuan and Yin, Fangcong and Singhal, Prasann and Wadhwa, Manya and Liu, Zeyu Leo and Sprague, Zayne and Namuduri, Ramya and Hu, Bodun and Rodriguez, Juan Diego and Peng, Puyuan and Durrett, Greg},
     booktitle = {Proceedings of the 39th Conference on Neural Information Processing Systems (NeurIPS), Datasets and Benchmarks Track},
     year = {2025},
   }

4. arXiv

   Linear Representation Transferability Hypothesis: Leveraging Small Models to Steer Large Models

   Femi Bello, Anubrata Das, Fanzhi Zeng, Fangcong Yin, and Liu Leqi

   arXiv preprint arXiv:2506.00653, 2025

   arXiv Bib

   @article{bello2025linear,
     title = {Linear Representation Transferability Hypothesis: Leveraging Small Models to Steer Large Models},
     author = {Bello, Femi and Das, Anubrata and Zeng, Fanzhi and Yin, Fangcong and Leqi, Liu},
     journal = {arXiv preprint arXiv:2506.00653},
     year = {2025},
   }

5. COLM

   LongProc: Benchmarking Long-Context Language Models on Long Procedural Generation

   Xi Ye, Fangcong Yin^*, Yinghui He^*, Joie Zhang^*, Howard Yen^*, Tianyu Gao, Greg Durrett, and Danqi Chen

   In Proceedings of the Conference on Language Modeling (COLM), 2025

   arXiv Bib

   @inproceedings{ye2025longproc,
     title = {LongProc: Benchmarking Long-Context Language Models on Long Procedural Generation},
     author = {Ye, Xi and Yin, Fangcong and He, Yinghui and Zhang, Joie and Yen, Howard and Gao, Tianyu and Durrett, Greg and Chen, Danqi},
     booktitle = {Proceedings of the Conference on Language Modeling (COLM)},
     year = {2025},
   }

6. ICML

   Understanding Synthetic Context Extension via Retrieval Heads

   Xinyu Zhao, Fangcong Yin, and Greg Durrett

   In Proceedings of the 42nd International Conference on Machine Learning (ICML), 2025

   arXiv Bib

   @inproceedings{zhao2025synthetic,
     title = {Understanding Synthetic Context Extension via Retrieval Heads},
     author = {Zhao, Xinyu and Yin, Fangcong and Durrett, Greg},
     booktitle = {Proceedings of the 42nd International Conference on Machine Learning (ICML)},
     year = {2025},
   }

7. ICLR

   To CoT or not to CoT? Chain-of-thought Helps Mainly on Math and Symbolic Reasoning

   Zayne Sprague, Fangcong Yin, Juan Diego Rodriguez, Dongwei Jiang, Manya Wadhwa, Prasann Singhal, Xinyu Zhao, Xi Ye, Kyle Mahowald, and Greg Durrett

   In Proceedings of the 13th International Conference on Learning Representations (ICLR), 2025

   arXiv Bib

   @inproceedings{sprague2025cot,
     title = {To CoT or not to CoT? Chain-of-thought Helps Mainly on Math and Symbolic Reasoning},
     author = {Sprague, Zayne and Yin, Fangcong and Rodriguez, Juan Diego and Jiang, Dongwei and Wadhwa, Manya and Singhal, Prasann and Zhao, Xinyu and Ye, Xi and Mahowald, Kyle and Durrett, Greg},
     booktitle = {Proceedings of the 13th International Conference on Learning Representations (ICLR)},
     year = {2025},
   }

2024

1. NeurIPS

   LoFiT: Localized Fine-tuning on LLM Representations

   Fangcong Yin, Xi Ye, and Greg Durrett

   In Proceedings of the 38th Conference on Neural Information Processing Systems (NeurIPS). Also presented at the Workshop on Foundation Model Interventions (Oral), NeurIPS 2024 , 2024

   arXiv Bib

   @inproceedings{yin2024lofit,
     title = {LoFiT: Localized Fine-tuning on LLM Representations},
     author = {Yin, Fangcong and Ye, Xi and Durrett, Greg},
     booktitle = {Proceedings of the 38th Conference on Neural Information Processing Systems (NeurIPS)},
     year = {2024},
   }

2023

1. EMNLP

   Linguistic Compression in Single-Sentence Human-Written Summaries

   Fangcong Yin and Marten van Schijndel

   In Findings of the Association for Computational Linguistics: EMNLP 2023, 2023

   Bib HTML

   @inproceedings{yin2023linguistic,
     title = {Linguistic Compression in Single-Sentence Human-Written Summaries},
     author = {Yin, Fangcong and {van Schijndel}, Marten},
     booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2023},
     year = {2023},
   }