Muhan Zhang

8

Papers

33

Total Citations

Papers (8)

PHYBench: Holistic Evaluation of Physical Perception and Reasoning in Large Language Models

NeurIPS 2025arXiv

Beyond Single-Task: Robust Multi-Task Length Generalization for LLMs

CLOVER: Cross-Layer Orthogonal Vectors Pruning

LoRASuite: Efficient LoRA Adaptation Across Large Language Model Upgrades

An Empirical Study of Realized GNN Expressiveness

Case-Based or Rule-Based: How Do Transformers Do the Math?

Graph As Point Set

OCN: Effectively Utilizing Higher-Order Common Neighbors for Better Link Prediction