郭司坤
PhD Candidate in Computer Science, University of Virginia
RealAI Lab, University of Virginia · Advisor: Prof. Aidong Zhang · Charlottesville, VA
Research areas: LLM reasoning, knowledge recruitment and in-context learning, inference-time learning, graph neural networks, AI for Science, and autonomous AI research systems.
Publications
Publications are listed in reverse chronological order. This page includes both first-author and collaborative work because the research agenda is broader than author position alone.
Sikun Guo, Di Wang, Xiaohan Fan, Albert Huang and Aidong Zhang
KDD 2026
Studies whether curiosity-oriented questioning can make research ideation with language models more exploratory, engine-agnostic, and practically useful. The method emphasizes question generation, filtering, and clarity scoring before ideas are handed to downstream ideation engines.
Sikun Guo, Amir Hassan Shariatmadari, Jiuqi Wang, Albert Huang, Stefan Bekiranov, Shangtong Zhang and Aidong Zhang
KDD 2026
Explores inference-time reinforcement learning as a way to optimize the quality of model-generated research ideas without relying only on larger base models. The project frames idea improvement as a trajectory with actions, rewards, and iterative policy updates at test time.
Amir Hassan Shariatmadari, Sikun Guo, Nathan C. Sheffield, Aidong Zhang and Kishlay Jha
IEEE ICDM 2025
ICDM 2025 Best Paper Award
HyHG models biomedical hypothesis generation with a temporal hypergraph view of evolving scientific concepts, allowing the system to capture higher-order relations that ordinary pairwise graphs miss. The paper received the ICDM 2025 Best Paper Award, making it a particularly visible result in the structured-scientific-discovery line of work.
Amir Hassan Shariatmadari, Alireza Jafari, Sikun Guo, Sneha Srinivasan, Nathan C. Sheffield, Aidong Zhang and Kishlay Jha
Bioinformatics 2025
ConceptDrift models biomedical hypothesis generation through the spatial, temporal, and semantic evolution of concepts. It represents a more structured scientific-modeling direction in the broader agenda, showing how graph-based views of evolving knowledge can support discovery.
Sikun Guo, Amir Hassan Shariatmadari, Peng Wang, Albert Huang and Aidong Zhang
Findings of EMNLP 2025
InfAL studies how adversarial learning at inference time can improve the quality of model-generated research ideas without changing the base model weights. The paper pushes research ideation beyond prompt engineering toward explicitly optimized test-time behavior.
Sikun Guo, Amir Hassan Shariatmadari, Guangzhi Xiong, Albert Huang, Myles Kim, Corey Williams, Stefan Bekiranov and Aidong Zhang
KDD 2025
IdeaBench introduces a benchmark and evaluation framework for LLM-based research idea generation. The core contribution is not only a dataset, but also a way to measure whether generated ideas are insightful, plausible, and meaningfully distinct rather than merely well phrased.
Guangzhi Xiong, Eric Xie, Corey Williams, Myles Kim, Amir Hassan Shariatmadari, Sikun Guo, Stefan Bekiranov and Aidong Zhang
IJCAI 2025
This paper focuses on the reliability of model-generated biomedical hypotheses, with particular attention to truthfulness and hallucination. It strengthens the agenda by asking not only whether models can generate hypotheses, but whether those hypotheses remain grounded and trustworthy.
Sikun Guo, Guangzhi Xiong and Aidong Zhang
SDM 2025
This paper studies how internal parametric knowledge and external knowledge sources should be balanced in scientific-discovery workflows. It connects directly to the broader question of how models should use what they already know instead of relying only on external retrieval.
Guangzhi Xiong, Eric Xie, Amir Hassan Shariatmadari, Sikun Guo, Stefan Bekiranov and Aidong Zhang
arXiv preprint
This preprint explores how knowledge grounding can improve scientific hypothesis generation with large language models, complementing later work on truthfulness, hallucination, and reliability in biomedical hypothesis generation.
Sikun Guo, Amir Hassan Shariatmadari, Guangzhi Xiong and Aidong Zhang
IEEE BigData 2024
A perspective on how foundation models can reshape scientific discovery workflows. The paper situates hypothesis generation, knowledge use, and human–AI collaboration within a larger AI-for-science agenda.
Guimin Dong, Mingyue Tang, Zhiyuan Wang, Jiechao Gao, Sikun Guo, Lihua Cai, Robert Gutierrez, Bradford Campbell, Laura E. Barnes and Mehdi Boukhechba
ACM Transactions on Sensor Networks
This survey organizes graph neural network research in IoT around graph modeling choices, node representations, edge semantics, and representative application settings. It reflects an earlier systems-and-graph-learning thread in my work that connects naturally to later interests in structured scientific knowledge.
Sikun Guo
CISCE 2020
An early survey on graph attention network-style methods, reflecting the graph-learning thread that preceded later work on robust GNNs and structured scientific knowledge.