Oral Sessions at CPAL 2026
A select number of papers from the CPAL 2026 Proceedings Track will be presented as oral presentations at the conference. The oral presentations are listed below, in their corresponding oral sessions.
Quick Links
Highlight Talks I
Time: Day 2 (Mar 24) – Tuesday – 3:30 PM to 4:00 PM
Matrix Sensing with Kernel Optimal Loss: Robustness and Optimization Landscape
Xinyuan Song, Ziye Ma
Keywords: Matrix sensing, kernel loss function, optimization
Teaching LLMs According to Their Aptitude: Adaptive Switching Between CoT and TIR for Mathematical Problem Solving
Xin Xu, Yan Xu, Tianhao Chen, Yuchen Yan, Chengwu Liu, Zaoyu Chen, Yufei Wang, Yichun Yin, Yasheng Wang, Qun Liu, Lu Yin
Keywords: Large Language Models, math QA, chain-of-thought, tool-integrated reasoning, fine-tuning
Sparse Mixture-of-Experts for Compositional Generalization: Empirical Evidence and Theoretical Foundations of Optimal Sparsity
Jinze Zhao, Peihao Wang, Junjie Yang, Ruisi Cai, Gaowen Liu, Jayanth Srinivasa, Ramana Rao Kompella, Yingbin Liang, Zhangyang Wang
Keywords: Compositional Generalization, Sparsity, Mixture of Experts
Highlight Talks II
Time: Day 3 (Mar 25) – Wednesday – 1:30 PM to 2:00 PM
From sparse recovery to plug-and-play priors, understanding trade-offs for stable recovery with generalized projected gradient descent
Ali Joundi, Yann Traonmilin, Jean-François Aujol
Keywords: Inverse Problems, Sparse Recovery, Plug-and-Play, Deep Prior, Optimization
Data-Efficient and Robust Trajectory Generation through Pathlet Dictionary Learning
yuanbo tang, Yan Tang, Zihui Zhao, Zixuan Zhang, Yang Li
Keywords: trajectory generative model, dictionary learning, sparse representation
Learning in the Null Space: Small Singular Values for Continual Learning
Cuong Anh Pham, Praneeth Vepakomma, Samuel Horváth
Keywords: continual learning, singular value decomposition, small singular values, null space
Highlight Talks III
Time: Day 4 (Mar 26) – Thursday – 1:30 PM to 2:00 PM
Analyzing and Mitigating Model Collapse in Reflow Methods
Huminhao Zhu, Fangyikang Wang, Tianyu Ding, Qing Qu, Zhihui Zhu
Keywords: Model Collapse, Self-training, Synthetic Data, Reflow, Rectified Flow
ROSE: Reordered SparseGPT for More Accurate One-Shot Large Language Models Pruning
Mingluo Su, Huan Wang
Keywords: Large language models, Unstructured pruning, Pruning order
What Scalable Second-Order Information Knows for Pruning at Initialization
Ivo Gollini Navarrete, Nicolas Mauricio Cuadrado, Martin Takáč, Samuel Horváth
Keywords: Pruning, Hessian, One-shot, Initialization, Hutchinson, Fisher