Paper-Conference

LIFT: A Novel Framework for Enhancing Long-Context Understanding of LLMs via Long Input Fine-Tuning

Y. Mao, Y. Xu, J. Li, F. Meng, H. Yang, Z. Zheng, X. Wang, M. Zhang

Relational In-Context Learning via Synthetic Pre-training with Structural Prior

Y. Wang, J. You, C. Shi, M. Zhang

SHINE: A Scalable In-Context Hypernetwork for Mapping Context to LoRA in a Single Pass

Y. Liu, X. Wang, Y. Mao, Y. Gelberg, H. Maron, M. Zhang

SubTokenTest: A Practical Benchmark for Real-World Sub-token Understanding

S. Hou, Y. Hu, M. Zhang

Towards a Mechanistic Understanding of Large Reasoning Models: A Survey of Training, Inference, and Failures

Y. Hu, J. Gu, R. Wang, Z. Yao, H. Peng, X. Wu, J. Chen, M. Zhang, L. Pan

VecMol: Vector-Field Representations for 3D Molecule Generation

Y. Hua, X. Peng, J. Ma, M. Zhang

TPLA: Tensor Parallel Latent Attention for Efficient Disaggregated Prefill & Decode Inference

X. Tang, F. Meng, P. Tang, Y. Wang, D. Yin, X. Sun, M. Zhang

Efficient Neural Common Neighbor for Temporal Graph Link Prediction

X. Zhang, Y. Wang, X. Wang, M. Zhang

FLORA: GNNs as Predictors of Agentic Workflow Performances

Y. Zhang, Y. Hou, B. Tang, M. Zhang, X. Dong, S. Chen

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

Length generalization, the ability to solve problems longer than those seen during training, remains a critical challenge for large …

Y. Hu, S. Kang, H. Yang, H. Xu, M. Zhang