newsitem: ICML 2026
Our Data and AI cluster is happy to present 14 contributions to the ICML 2026 conference:
Main track
MEAL: A Benchmark for Continual Multi-Agent Reinforcement Learning
Tristan Tomilin* (TU/e - DAI), Luka van den Boogaard* (TU/e - DAI), Samuel Garcin, Constantin Ruhdorfer, Bram Grooten* (TU/e - DAI), Fabrice Kusters* (TU/e - DAI), Yali Du, Andreas Bulling, Mykola Pechenizkiy* (TU/e - DAI), Meng Fang* (TU/e - DAI)
Self-Evolving LLM Agents under Offline Data Support
Yudi Zhang* (TU/e - DAI), Meng Fang* (TU/e - DAI), Zhenfang Chen, Mykola Pechenizkiy* (TU/e - DAI)
When Data Is Scarce: Scaling Sparse Language Models with Repeated Training
Boqian Wu, Qiao Xiao* (TU/e - DAI), Patrik Okanovic, Tomasz Sternal, Maurice van Keulen, Mykola Pechenizkiy* (TU/e - DAI), Elena Mocanu, Torsten Hoefler, Decebal Constantin Mocanu
Memory-Efficient LLMs Training with Dynamic Sparsity: From Stability to Practical Scaling
Qiao Xiao* (TU/e - DAI), Boqian Wu, Patrik Okanovic, Tomasz Sternal, Maurice van Keulen, Elena Mocanu, Mykola Pechenizkiy* (TU/e - DAI), Decebal Constantin Mocanu, Torsten Hoefler
STFlow: Data-Coupled Flow Matching for Geometric Trajectory Simulation
Kiet Bennema ten Brinke* (TU/e - DAI), Koen Minartz* (TU/e - DAI), Vlado Menkovski* (TU/e - DAI)
GradientStabilizer: Fix the Norm, Not the Gradient
Tianjin Huang* (TU/e - DAI), Zhangyang Wang, Haotian Hu, Zhenyu Zhang, Gaojie Jin, Xiang Li, Li Shen, Jiaxing Shang, Tianlong Chen, Ke Li, Lu Liu, Qingsong Wen, Shiwei Liu
OTora: A Unified Red Teaming Framework for Reasoning-Level Denial-of-Service in LLM Agents
Xinyu Li, Ronghui Mu, Lin Li, Tianjin Huang* (TU/e - DAI), Gaojie Jin
Margin-Adaptive Confidence Ranking for Reliable LLM Judgement
Gaojie Jin, Yong Tao, Lijia Yu, Tianjin Huang* (TU/e - DAI)
Towards Uniformity and Alignment for Multimodal Representation Learning
Wenzhe Yin, Pan Zhou, Zehao Xiao, Jie Liu* (TU/e - DAI), Shujian Yu, Jan-Jakob Sonke, Stratis Gavves
Tracks and Workshops
On the Learning Dynamics of Label-Noise Memorization in ReLU MLPs — MEMFM workshop
Yannis Kaltampanidis* (TU/e - DAI), Mykola Pechenizkiy* (TU/e - DAI), Hannah Pinson* (TU/e - DAI)
Reconstructing Training Images from Foundation Model Parameters in the Healthcare Domain: Privacy Risks and Defences — MEMFM workshop
Nassos Glykos* (TU/e - DAI), Yannis Kaltampanidis* (TU/e - DAI), Mykola Pechenizkiy* (TU/e - DAI), Hannah Pinson* (TU/e - DAI)
Position: AI Should Verify, Not Judge, Scientific Work — AI for Science workshop
Prabhant Singh* (TU/e - DAI), Thanh Gia Hieu Khuong, Vlasta Sikimić, Benedictus Kent Rachmat, Kola Ayonrinde, Ihsan Ullah, Christina Lioma, Luis Oala, Kevin Qinghong Lin, Lele Cao, Neil F. Abernethy, Hilde Weerts* (TU/e - DAI), Joaquin Vanschoren* (TU/e - DAI)
Is Our Benchmark Enough? An Analysis of Continual Learning for MLLMs — CATS workshop
Van-Tuan Tran, Shruthi Gowda* (TU/e - DAI), Merim Dzaferagic, Marco Ruffini
COMRAD: A Benchmark for Embodied Cooperative Multi-Agent Reinforcement Learning — New Frontiers in Game-Theoretic Learning workshop
Khoi Nguyen, Dimitar Z Zhekov, Tristan Tomilin* (TU/e - DAI)