Yiling Ma

About me

I am a master’s student in the Department of Biostatistics at Yale University, where I work with Prof. Arman Cohan and Dr. Manasi Patwardhan on LLMs for research at the Yale NLP Lab. I will complete my master’s thesis under the supervision of Prof. Qingyu Chen.

I have also worked with VLLab @ UC Merced, AI4Biomed Lab, Gerstein Lab, OIST CSE, the University of Florida, Saudi Aramco, New York University Shanghai, Zhejiang Lab, and Wisbot (startup). I appreciate the opportunity to collaborate with distinguished professors and researchers from these institutions.

Currently: finishing my thesis and preparing PhD applications.

Research interests

My research focuses on building reliable AI systems that can reason over evidence, integrate information across modalities, and improve through feedback and interaction. I am especially interested in language and multimodal models for scientific and high-stakes decision-making, where models must be not only capable, but also grounded, transparent, and adaptive.

1. Trustworthy LLMs & Reasoning:→ show/hide interest in faithful, evidence-grounded language model reasoning.

Question: How can we make model outputs not only stronger, but reliably grounded in the right evidence, scope, and reasoning process?

I am interested in developing language models that can reason over complex evidence, recognize uncertainty or ambiguity, and produce outputs that are faithful to the underlying information rather than merely plausible.

2. Multimodal AI for X:→ show/hide interest in connecting heterogeneous evidence across modalities.

Question: How can AI systems connect language, vision, temporal signals, and structured data to support reasoning in complex real-world domains?

I am interested in multimodal models that integrate heterogeneous evidence, such as text, images, videos, and longitudinal signals, especially in scientific and medical settings where important information is distributed across different data sources.

3. Self-Improvement:→ show/hide interest in feedback, memory, and iterative refinement.

Question: How can AI systems learn to improve their own reasoning, feedback, and decision-making over time?

I am interested in models that can benefit from feedback, memory, external knowledge, and iterative refinement, moving from one-shot generation toward systems that can diagnose their own weaknesses and become more reliable through interaction.

Selected Publications

When Evidence Vanishes: Distilling Visual-Evidence Dependency in Medical Vision-Language Models

Qizhen Lan^*, Yiling Ma^*, Ruiqi Chen, Mengchen Fan, Xi Xiao, Lin Zhao, Liang He, Baocheng Geng, Pulin Che, Tianyang Wang, Xiaoqian Jiang

* Equal contribution (co-first authors).

arXiv

IdeaAMBIG: Benchmarking Implementation-Critical Gaps in Research-Idea Specifications

Yiling Ma, Yilun Zhao, Sihong Wu, Manasi Patwardhan, Arman Cohan

arXiv

ActReview: Rebuttal-Guided Training Data and Rubric Rewards for Actionable Peer Review Generation

Yiling Ma, Yilun Zhao, Sihong Wu, Ziyu Chen, Manasi Patwardhan, Arman Cohan

arXiv

Beyond the Embedding Bottleneck: Adaptive Retrieval-Augmented 3D CT Report Generation

Renjie Liang, Yiling Ma, Yang Xing, Zhengkang Fan, Jinqian Pan, Chengkun Sun, Li Li, Kuang Gong, Jie Xu

arXiv

RbtAct: Rebuttal as Supervision for Actionable Review Feedback Generation

Sihong Wu, Yiling Ma, Yilun Zhao, Tiansheng Hu, Owen Jiang, Manasi Patwardhan, Arman Cohan

ACL 2026

Paper

Can AI Be a Good Peer Reviewer? A Survey of Peer Review Process, Evaluation, and the Future

Sihong Wu, Owen Jiang, Yilun Zhao, Tiansheng Hu, Yiling Ma, Kaiyan Zhang, Manasi Patwardhan, Arman Cohan

ACL 2026

Paper

A Systematic Review of Deep Learning and Machine Learning Applications in Longitudinal Multimodal Clinical Data

Jinqian Pan, Tienyu Chang, Mengxian Lyu, Weimin Meng, Qingyu Wang, Yiling Ma, Ziyi Chen, Xiaohan Li, Chengkun Sun, Renjie Liang, Jennifer Fishe, Jie Xu

Journal of Healthcare Informatics Research, 2026

Paper

Bridging Gene Regulatory Networks and Causal Representation Learning in Single-Cell Genomics Data

Vincenzo Lagani, Giorgi Sokhadze, Liliia Nigmetzianova, Robert Lehmann, Yiling Ma, Sumeer Khan, Xabier Martínez de Morentin, Narsis Kiani, Mikel Hernaez, Alexander A Lukyanov, Jesper Tegnér, David Gomez-Cabrero