Time	Event	Details
09:00–09:10	Welcome	Chair: Tiansi Dong
09:10–09:50	Keynote 1: Roberto Navigli (Sapienza University of Rome) – Exploring Semantics in the Age of Large Language Models	Chair: Tiansi Dong
09:50–10:30	Keynote 2: Regina Zhang (NTU) - AI-Powered Graph Representation Learning for Robust and Efficient Urban, Social and Biological Science	Chair: Kang Liu
10:30–11:00	Morning coffee break
11:00–11:45	Keynote 3: Liu Kang (Chinese Academy of Sciences) - Shuttle between Symbolic Knowledge and Neural Parameters	Chair: Tiansi Dong
11:45–12:30	Keynote 4: Minghui Dong (A*STAR I²R) - Beyond Fluency: Accountability and Declined Answers in Safety-Critical LLMs	Chair: Kang Liu
12:30–14:00	Lunch break
14:00–14:45	Keynote 5: Ruihong Huang (Texas A&M University) - Discourse Structure Guided NLP Models for Fine-grained Media Bias Analysis	Chair: Tiansi Dong
14:45–15:30	Keynote 6: Zheng Wang (Huawei Singapore Research Center) - openJiuwen for AI Agent Practice: Advanced Knowledge Retrieval and Workflow Planning	Chair: Ying Jiao
15:20–16:00	Afternoon coffee break
16:00–16:30	Oral Session	Chair: Ying Jiao
	Oral 1: Stabilizing Reinforcement Learning for Honesty Alignment in Language Models on Deductive Reasoning	Jiarui Liu (CMU), Kaustubh Dhole (Emory University), Yingheng Wang (Cornell), Haoyang Wen (Amazon), Sarah Zhang (Amazon), Haitao Mao (Michigan State University), Gaotang Li (UIUC), Neeraj Varshney (Amazon), Jingguo Liu (Amazon), Xiaoman Pan (Amazon)
	Oral 2: Learning to Compose Symbolic Jazz	Arthur Wayne (University of Pennsylvania), Jonathan Bachrach
	Oral 3: Neurosymbolic LoRA: Why and When to Tune Weights vs. Rewrite Prompts	Kevin Wang (UT Austin), Neel P. Bhatt (UT Austin), Cong Liu (Collins Aerospace), Junbo Li (UT Austin), Runjin Chen (UT Austin), Yihan Xi (UT Austin), Timothy Barclay (Collins Aerospace), Alvaro Velasquez (DARPA, University of Colorado at Boulder), ufuk topcu (UT Austin), Zhangyang Wang (UT Austin)
16:30–17:30	Poster Session
	GeoSteer: Faithful Chain-of-Thought Steering via Latent Manifold Gradients	Kentaro Kazama (Mitsubishi Electric Corporation), Daiki Shirafuji (Mitsubishi Electric Corporation), Tatsuhiko Saito (Mitsubishi Electric Corporation)
	N2N-GQA: Noise-to-Narrative for Graph-based Table-Text Question Answering using LLMs	Aravindh Annamalai (Comcast Corporation)
	Improving Symbolic Translation of Language Models for Logical Reasoning	Ramya Keerthy Thatikonda (Monash University), Jiuzhou Han (Monash University), Wray Buntine (VinUniversity), Ehsan Shareghi (Monash University)
	The Inside of Reasoning: Uncovering Algebraic Structure in Bilinear MLPs	Kaustubh Sharma (Indian Institute of Technology Roorkee), Aditya Chauhan (Indian Institute of Technology Roorkee)
	On the Limitations of Rank-One Model Editing for Multi-Hop Reasoning in Transformers	Zhiyuan He (UCL), Binghan Chen (UCL), Tianxiang Xiong (UCL), Ziyang Sun (UCL), Mozhao Zhu (UCL), Xi Chen (UCL)
	Achieving Near-Specialist Performance with Vision-Language-Action Generalist Models Using a Novel Neuro-symbolic Task Configuration Framework	Anindya Jana (Jadavpur University), Snehasis Banerjee (IIIT Guwahati, Tata Consultancy Services), Arup Kumar Sadhu (Tata Consultancy Services), Ranjan Dasgupta (Tata Consultancy Services)
	Gist Meets Graph: A Dual-Trace Neuro-Symbolic Reasoner for Knowledge Graph Question Answering	Md Muntaqim Meherab (Daffodil International University)
	From Sentences to Proof Trees: Leveraging Language Models for Structured Reasoning	Aayushee Gupta (International Institute of Information Technology, Bangalore), G. Srinivasaraghavan
	Do Reasoning Models Ask Better Questions? A Formal Information-Theoretic Analysis on Multi-Turn LLM Games	Daniel Machado Pedrozo (Universidade Federal de Goiás), Telma Woerle de Lima Soares (Universidade Federal de Goiás), Bryan Lincoln Marques de Oliveira (Universidade Federal de Goiás)
	Tri-CoT: Multi-Perspective Chain-of-Thought for Implicit Hate Speech Detection	Yucheng Lin (University of New South Wales), Jiaojiao Jiang (University of New South Wales)
	ConceptGuard: Neuro-Symbolic Safety Guardrails via Sparse Interpretable Jailbreak Concepts	Darpan Aswal (Université Paris-Saclay), Celine Hudelot (CentraleSupelec)
17:30–17:45	Closing Session & Best Paper Awards	Chair: Tiansi Dong