Preprints

•   Discovering a Shared Logical Subspace: Steering LLM Logical Reasoning via Alignment of Natural-Language and Symbolic View

  (LLM logical reasoning + neural-symbolic reasoning)

      Feihao Fang, My Thai, Yuanyuan Lei

      in submission to the Association for Computational Linguistics (ACL) Rolling Review [pdf]

      keywords: LLM logical reasoning, neural-symbolic reasoning, knowledge steering

•   Where Reasoning Breaks: Logic-Aware Path Selection by Controlling Logical Connectives in LLMs Reasoning Chains

  (LLM logical reasoning)

      SeungHyun Park, Yuanyuan Lei

      in submission to the Association for Computational Linguistics (ACL) Rolling Review [pdf]

      keywords: LLM logical reasoning, reasoning fragility, logical relations, inference-time intervention

•   Large Language Models in Resolving Contextual Knowledge Conflicts (LLM + knowledge conflict)

      Xinye Yang, Zhenyang Liu, Ruisi Li, Yuanyuan Lei

      in submission to the Association for Computational Linguistics (ACL) Rolling Review [pdf]

      keywords: LLM, knowledge conflict, context conflict, critical thinking

•   Knowledge Vector of Logical Reasoning in Large Language Models (LLM + knowledge + logical reasoning)

      Zixuan Wang, Yuanyuan Lei

      in submission to the Association for Computational Linguistics (ACL) Rolling Review [pdf]

      keywords: LLM logical reasoning, knowledge representation, mechanism interpretability

•   Analyzing Intrinsic Self-Correction in Video Reasoning: Inference-Time Behavior and Training Learnability

  (VLM video reasoning + self-correction)

      Ye Zhang, Yuanyuan Lei

      in submission to the Association for Computational Linguistics (ACL) Rolling Review [pdf]

      keywords: video-language reasoning, vision-language models, self-correction

•   Meta-Planning for Large Language Models Planning (LLM reasoning and planning)

      Yuanyuan Lei, Fei Liu, Hongyuan Mei, Chinmay Kulkarni

      in submission to the Association for Computational Linguistics (ACL) Rolling Review [pdf]

      keywords: LLM reasoning and planning, LLM agent, multi-stage reasoning, meta-planning

Conference and Journal Publications

•   Multi-document Summarization through Multi-document Event Relation Graph Reasoning in LLMs

  (LLM + summarization + knowledge reasoning)

      Yuanyuan Lei, Ruihong Huang

      Proceedings of the 2025 Annual Meetings of the Association for Computational Linguistics (ACL 2025) [pdf]

      keywords: multi-document summarization, multi-document knowledge reasoning, knowledge-augmented LLM

•   Boosting Logical Fallacy Reasoning in LLMs via Logical Structure Tree (LLM + logical reasoning + knowledge reasoning)

      Yuanyuan Lei, Ruihong Huang

      Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024) [pdf]

      keywords: logical reasoning, logical fallacy reasoning, logical structure, knowledge-augmented LLM

•   Evaluating Large Language Models in Making Moral Judgments

      Divij Bajaj, Yuanyuan Lei, Jonathan Tong, Ruihong Huang

      Findings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024) [pdf]

      keywords: LLM moral foundations, moral evaluation and analysis in LLM

•   Sentence-level Media Bias Analysis with Event Relation Graph

  Yuanyuan Lei, Ruihong Huang

  Proceedings of the 2024 North America Chapter of the Association for Computational Linguistics (NAACL 2024) [pdf]

  keywords: information framing detection, event-level narrative structure, knowledge extraction and reasoning

•   Polarity Calibration for Opinion Summarization (LLM + summarization + knowledge alignment)

  Yuanyuan Lei, Kaiqiang Song, Sangwoo Cho, Xiaoyang Wang, Ruihong Huang, Dong Yu

  Proceedings of the 2024 North America Chapter of the Association for Computational Linguistics (NAACL 2024) [pdf]

  keywords: multi-document summarization, controllable text generation, knowledge alignment, reinforcement learning

•   EMONA: Event-level Moral Opinions in News Articles

  Yuanyuan Lei, Md Messal Monem Miah, Ayesha Qamar, Sai Ramana Reddy, Jonathan Tong, Haotian Xu, Ruihong Huang

  Proceedings of the 2024 North America Chapter of the Association for Computational Linguistics (NAACL 2024) [pdf] [data]

  keywords: opinions mining, moral opinions, moral framing, benchmarking

•   Discourse Structures Guided Fine-grained Propaganda Identification

  Yuanyuan Lei, Ruihong Huang

  Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023) [pdf]

  keywords: propaganda detection, inter-sentence discourse relations, knowledge reasoning, knowledge distillation

•   Identifying Conspiracy Theories News based on Event Relation Graph

  Yuanyuan Lei, Ruihong Huang

  Findings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023) [pdf]

  keywords: conspiracy theories detection, event-level logical relations, knowledge extraction and reasoning

•   Audio-Visual Emotion Recognition with Preference Learning based on Intended and Multi-modal Perceived Labels

  Yuanyuan Lei, Houwei Cao

  IEEE Transactions on Affective Computing (IEEE Transactions), 2023, regular paper, impact factor: 9.8 [pdf]

  keywords: multi-modality learning, audio-visual content understanding, learning to rank, preference learning

•   Sentence-level Media Bias Analysis Informed by Discourse Structures

  Yuanyuan Lei, Ruihong Huang, Lu Wang, Nick Beauchamp

  Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022) [pdf]

  keywords: information framing detection, discourse structure, knowledge reasoning, knowledge distillation

•   Few-Shot (Dis)Agreement Identification in Online Discussions with Regularized and Augmented Meta-Learning

  Yuanyuan Lei, Ruihong Huang

  Findings of the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP 2022) [pdf] [data]

  keywords: argument mining, meta-learning, domain adaptation

•   Cocktail Edge Caching: Ride Dynamic Trends of Content Popularity With Ensemble Learning

  Tongyu Zong, Chen Li, Yuanyuan Lei, Guangyu Li, Houwei Cao and Yong Liu

  IEEE/ACM Transactions on Networking (IEEE Transactions), 2022, regular paper [pdf]

  keywords: multi-modality learning, deep reinforcement learning