22.04.2026
MCML at ICLR 2026
36 Accepted Papers (33 Main, and 3 Workshops)
14th International Conference on Learning Representations, Rio de Janeiro, Brazil, Apr 23-27, 2026
We are happy to announce that MCML researchers have contributed a total of 36 papers to ICLR 2026: 33 Main, and 3 Workshop papers. Congrats to our researchers!
Main Track (33 papers)
E. Araya • M. Datres • G. Kutyniok
Random Spiking Neural Networks are Stable and Spectrally Simple.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
Random Spiking Neural Networks are Stable and Spectrally Simple.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
J. von Berg • A. Fono • M. Datres • S. Maskey • G. Kutyniok
The Price of Robustness: Stable Classifiers Need Overparameterization.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
The Price of Robustness: Stable Classifiers Need Overparameterization.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
S. Ball • G. Gluch • S. Goldwasser • F. Kreuter • O. Reingold • G. N. Rothblum
On the Impossibility of Separating Intelligence from Judgment: The Computational Intractability of Filtering for AI Alignment.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
On the Impossibility of Separating Intelligence from Judgment: The Computational Intractability of Filtering for AI Alignment.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
A. Baumann • R. Li • M. Klasson • S. Mentu • S. Karthik • Z. Akata • A. Solin • M. Trapp
Post-hoc Probabilistic Vision-Language Models.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Post-hoc Probabilistic Vision-Language Models.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
W. Chen • C. Zheng • G. Zhang • A. Vedaldi • D. Cremers
NOVA3R: Non-pixel-aligned Visual Transformer for Amodal 3D Reconstruction.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
NOVA3R: Non-pixel-aligned Visual Transformer for Amodal 3D Reconstruction.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
C. Datar • T. Kapoor • A. Chandra • Q. Sun • E. L. Bolager • I. Burak • A. Veselovska • M. Fornasier • F. Dietrich
Fast training of accurate physics-informed neural networks without gradient descent.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Fast training of accurate physics-informed neural networks without gradient descent.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
M. Fayyaz • A. Modarressi • H. Deilamsalehy • F. Dernoncourt • R. Rossi • T. Bui • H. Schütze • N. Peng
Steering MoE LLMs via Expert (De)Activation.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv GitHub
Steering MoE LLMs via Expert (De)Activation.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv GitHub
L. Girrbach • S. Alaniz • G. Smith • T. Darrell • Z. Akata
Person-Centric Annotations of LAION-400M: Auditing Bias and Its Transfer to Models.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Person-Centric Annotations of LAION-400M: Auditing Bias and Its Transfer to Models.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
E. Guha • R. Marten • S. Keh • N. Raoof • G. Smyrnis • H. Bansal • M. Nezhurina • J. Mercat • T. Vu • Z. Sprague • A. Suvarna • B. Feuer • L. Chen • Z. Khan • E. Frankel • S. Grover • C. Choi • N. Muennighoff • S. Su • W. Zhao • J. Yang • S. Pimpalgaonkar • K. Sharma • C. C.-J. Ji • Y. Deng • S. Pratt • V. Ramanujan • J. Saad-Falcon • J. Li • A. Dave • A. Albalak • K. Arora • B. Wulfe • C. Hegde • G. Durrett • S. Oh • M. Bansal • S. Gabriel • A. Grover • K.-W. Chang • V. Shankar • A. Gokaslan • M. A. Merrill • T. Hashimoto • Y. Choi • J. Jitsev • R. Heckel • M. Sathiamoorthy • A. G. Dimakis • L. Schmidt
OpenThoughts: Data Recipes for Reasoning Models.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv URL
OpenThoughts: Data Recipes for Reasoning Models.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv URL
M. Gui • J. Schusterbauer • T. Phan • F. Krause • J. Susskind • M. A. Bautista • B. Ommer
Adapting Self-Supervised Representations as a Latent Space for Efficient Generation.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Adapting Self-Supervised Representations as a Latent Space for Efficient Generation.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
K. Hess • D. Frauen • V. Melnychuk • S. Feuerriegel
IGC-Net for conditional average potential outcome estimation over time.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
IGC-Net for conditional average potential outcome estimation over time.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
K. Heß • D. Frauen • V. Melnychuk • S. Feuerriegel
Efficient and Sharp Off-Policy Learning under Unobserved Confounding.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Efficient and Sharp Off-Policy Learning under Unobserved Confounding.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
K. Heß • D. Frauen • M. van der Schaar • S. Feuerriegel
Overlap-weighted orthogonal meta-learner for treatment effect estimation over time.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Overlap-weighted orthogonal meta-learner for treatment effect estimation over time.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
P. Hofman • T. Löhr • M. Muschalik • Y. Sale • E. Hüllermeier
Efficient Credal Prediction through Decalibration.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
Efficient Credal Prediction through Decalibration.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
E. Javurek • V. Melnychuk • J. Schweisthal • K. Heß • D. Frauen • S. Feuerriegel
An Orthogonal Learner for Individualized Outcomes in Markov Decision Processes.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
An Orthogonal Learner for Individualized Outcomes in Markov Decision Processes.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
J. Lan • Z. Liu • U. Schlegel • R. Zhao • Y. Liu • H. Schütze • M. A. Hedderich • T. Seidl
Human Uncertainty-Aware Data Selection and Automatic Labeling in Visual Question Answering.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Human Uncertainty-Aware Data Selection and Automatic Labeling in Visual Question Answering.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
V. Melnychuk • S. Feuerriegel
GDR-learners: Orthogonal Learning of Generative Models for Potential Outcomes.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
GDR-learners: Orthogonal Learning of Generative Models for Potential Outcomes.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Y. Ma • D. Frauen • E. Javurek • S. Feuerriegel
Foundation Models for Causal Inference via Prior-Data Fitted Networks.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Foundation Models for Causal Inference via Prior-Data Fitted Networks.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
V. Melnychuk • D. Frauen • J. Schweisthal • S. Feuerriegel
Overlap-Adaptive Regularization for Conditional Average Treatment Effect Estimation.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Overlap-Adaptive Regularization for Conditional Average Treatment Effect Estimation.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
R.-A. Matişan • V. T. Hu • G. Bartosh • B. Ommer • C. G. M. Snoek • M. Welling • J.-W. van de Meent • M. M. Derakhshani • F. Eijkelboom
Purrception: Variational Flow Matching for Vector-Quantized Image Generation.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Purrception: Variational Flow Matching for Vector-Quantized Image Generation.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
S. Maskey • R. Paolino • F. Jogl • G. Kutyniok • J. Lutzeyer
Graph Representational Learning: When Does More Expressivity Hurt Generalization?
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Graph Representational Learning: When Does More Expressivity Hurt Generalization?
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
J. McGinnis • S. Shit • F. A. Hölzl • P. Friedrich • P. Büschl • V. Sideri-Lampretsa • M. Mühlau • P. C. Cattin • B. Menze • D. Rückert • B. Wiestler
Beyond Uniformity: Regularizing Implicit Neural Representations through a Lipschitz Lens.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL GitHub
Beyond Uniformity: Regularizing Implicit Neural Representations through a Lipschitz Lens.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL GitHub
P. Mondorf • S. Zhou • M. Riedler • B. Plank
Compositional-ARC: Assessing Systematic Generalization in Abstract Spatial Reasoning.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Compositional-ARC: Assessing Systematic Generalization in Abstract Spatial Reasoning.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
A. Rahma • C. Datar • A. Cukarska • F. Dietrich
Rapid training of Hamiltonian graph networks without gradient descent.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Rapid training of Hamiltonian graph networks without gradient descent.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
T. Rochussen • V. Fortuin
Amortising Inference and Meta-Learning Priors in Neural Networks.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Amortising Inference and Meta-Learning Priors in Neural Networks.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
A. von Recum • L. Girrbach • Z. Akata
Are Reasoning LLMs Robust to Interventions on Their Chain-of-Thought?
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Are Reasoning LLMs Robust to Interventions on Their Chain-of-Thought?
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
M. Seleznova • H.-H. Chou • C. M. Verdun • G. Kutyniok
GradPCA: Leveraging NTK Alignment for Reliable Out-of-Distribution Detection.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
GradPCA: Leveraging NTK Alignment for Reliable Out-of-Distribution Detection.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
J. Tong • Y. Fan • A. Zhao • Y. Ma • X. Shen
StreamingThinker: Large Language Models Can Think While Reading.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL GitHub
StreamingThinker: Large Language Models Can Think While Reading.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL GitHub
X. Wang • N. Joshi • B. Plank • R. Angell • H. He
Is It Thinking or Cheating? Detecting Implicit Reward Hacking by Measuring Reasoning Effort.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Is It Thinking or Cheating? Detecting Implicit Reward Hacking by Measuring Reasoning Effort.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
K. Wang • T. Klug • S. Ruschke • J. Kirschke • R. Heckel
Reliable Evaluation of MRI Motion Correction: Dataset and Insights.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Reliable Evaluation of MRI Motion Correction: Dataset and Insights.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Y. Wu • Y. Wang • Z. Ye • T. Du • S. Jegelka • Y. Wang
When More is Less: Understanding Chain-of-Thought Length in LLMs.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
When More is Less: Understanding Chain-of-Thought Length in LLMs.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
E. Zaradoukas • B. Prenkaj • G. Kasneci
Reinforcement Unlearning via Group Relative Policy Optimization.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Reinforcement Unlearning via Group Relative Policy Optimization.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Y. Zhang • S. Tang • Z. Li • Z. Han • V. Tresp
WebArbiter: A Principle-Guided Reasoning Process Reward Model for Web Agents.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
WebArbiter: A Principle-Guided Reasoning Process Reward Model for Web Agents.
ICLR 2026 - 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. arXiv
Workshops (3 papers)
E. Garces Arias
A Credal-Set Perspective on Task-Induced Distributional Drift in Text Generation.
CAO @ICLR 2026 - Workshop Catch, Adapt, and Operate: Monitoring ML Models Under Drift at the 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
A Credal-Set Perspective on Task-Induced Distributional Drift in Text Generation.
CAO @ICLR 2026 - Workshop Catch, Adapt, and Operate: Monitoring ML Models Under Drift at the 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
T. Höppe • S. Bauer • Q. Liu • A. Dittadi • K. Neklyudov
On Closed-Form Couplings.
GRaM @ICLR 2026 - Workshop on Geometry-grounded Representation Learning and Generative Modeling at the 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
On Closed-Form Couplings.
GRaM @ICLR 2026 - Workshop on Geometry-grounded Representation Learning and Generative Modeling at the 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
M. Morik • L. Ciernik • L. Thede • L. Eyring • S. Nakajima • Z. Akata • L. Muttenthaler
Revealing Task-Dependent Layer Relevance via Attentive Multi-Layer Fusion.
Sci4DL @ICLR 2026 - Workshop on Scientific Methods for Understanding Deep Learning at the 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
Revealing Task-Dependent Layer Relevance via Attentive Multi-Layer Fusion.
Sci4DL @ICLR 2026 - Workshop on Scientific Methods for Understanding Deep Learning at the 14th International Conference on Learning Representations. Rio de Janeiro, Brazil, Apr 23-27, 2026. To be published. Preprint available. URL
Related
20.04.2026
Research Stay at École Polytechnique
Viktoria Ehm joined a research stay at École Polytechnique via MCML AI X-Change, working on 3D shape analysis and LLM-based methods.
