Bo Xiong
Postdoc @ Stanford University

I am a Postdoctoral Scholar in AI at Stanford University, under the supervision of Prof. Mark Musen. Prior to that, I received my Ph.D. (with summa cum laude) in computer science at University of Stuttgart, Germany and the Intl. Max Plank Research School for Intelligent Systems (IMPRS-IS), under the supervision of Prof. Steffen Staab. My PhD was funded by the Marie Skłodowska-Curie PhD Fellowship. I was also an associate member of the TrustAGI Lab advised by Prof. Shirui Pan. I have published 20+ papers and/or served as PC in premier AI conferences such as NeurIPS, ICLR, KDD, ACL, WWW, EMNLP, NAACL, SIGIR, AAAI, ECAI, ISWC, etc., and received the Best Student Paper Award of ISWC’22.
I am currently developing neuro-symbolic and knowledge-intensive foundation models, which are capable of learning, reasoning, and adapting into diverse tasks. I aim to bridge the gap between symbolic structures (e.g., knowledge graphs, biomedical ontology) and deep learning (e.g., LLMs, multi-modal models), and build reliable and interpretable AI systems allowing for adapting into applications in biomedicine and healthcare.
In my PhD dissertation (PDF), I studied a new paradigm of representation learning that captures both logical structures and similarity, bridging two historically distinct approaches in AI: connectionism and symbolism.
Recent News
Feb 2025 | One paper was accepted by IEEE TPAMI (IF 20.8) |
---|---|
Jan 2025 | One paper was accepted by ICLR’25 |
Jan 2025 | Two papers (1 main, 1 short) were accepted by WWW’25 |
Jan 2025 | One long paper was accepted by NAACL’25 |
Dec 2024 | I started a Postdoc position at Stanford University |
Sep 2024 | Three finding papers got accepted by EMNLP’24 |
Jul 2024 | I successfully defended my Ph.D. with summa cum laude (highest distinction) |
Selected Publications
-
ICLRFrom Tokens to Lattices: Emergent Lattice Structure in Language ModelsInternational Conference on Learning Representation 2025
-
WWWMaSH: Maximal Separating Poincaré Hyperplanes for Hierarchical Imbalanced LearningShort Paper Track, ACM The Web Conference 2025
-
WWWDAGE: DAG Query Answering via Relational Combinator with Logical ConstraintsThe Web Conference 2025
-
Preprint