03.07.2026
MCML at ICML 2026
76 Accepted Papers (64 Main, and 12 Workshops)
43rd International Conference on Machine Learning, Seoul, South Korea, Jul 06-11, 2026
We are happy to announce that MCML researchers have contributed a total of 76 papers to ICML 2026: 64 Main, and 12 Workshop papers. Congrats to our researchers!
Main Track (64 papers)
H. Arno • D. Frauen • E. Javurek • T. Demeester • S. Feuerriegel
Rank-Learner: Orthogonal Ranking of Treatment Effects.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Rank-Learner: Orthogonal Ranking of Treatment Effects.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
E. Antipov • A. Palma • L. Consoli • S. Günnemann • A. Dittadi • F. J. Theis
Flow-Based Density Ratio Estimation for Intractable Distributions with Applications in Genomics.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Flow-Based Density Ratio Estimation for Intractable Distributions with Applications in Genomics.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
S. Ball • S. Allmendinger • N. Kühl • F. Kreuter
Reading Between the Tokens: Improving Preference Predictions through Mechanistic Forecasting.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Reading Between the Tokens: Improving Preference Predictions through Mechanistic Forecasting.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
J. Bi • A. Aniri • M. Yang • X. Zhou • W. Huang • S. Yan • Y. Wang • Z. Cao • M. Färber • X. Xiao • V. Tresp • Y. Ma
EchoRL: Reinforcement Learning via Rollout Echoing.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
EchoRL: Reinforcement Learning via Rollout Echoing.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
J. Becktepe • A. Franz • N. Thuerey • S. Peitz
Plug-and-Play Benchmarking of Reinforcement Learning Algorithms for Large-Scale Flow Control.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Plug-and-Play Benchmarking of Reinforcement Learning Algorithms for Large-Scale Flow Control.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
S. Ball • P. Hackemann
Position: The Alignment Community is Unintentionally Building a Censor’s Toolkit.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Oral Presentation. To be published. Preprint available. GitHub
Position: The Alignment Community is Unintentionally Building a Censor’s Toolkit.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Oral Presentation. To be published. Preprint available. GitHub
S. Ball • A. Haupt
Don't Walk the Line: Boundary Guidance for Filtered Generation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. arXiv
Don't Walk the Line: Boundary Guidance for Filtered Generation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. arXiv
V. Bürgin • D. Herbst • Y.-W. E. Lin • S. Jegelka
Beyond Structural Symmetries: Linear Mode Connectivity via Neuron Identifiability.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Beyond Structural Symmetries: Linear Mode Connectivity via Neuron Identifiability.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
M. Brockschmidt • M. Schröder • S. Feuerriegel
SurvDiff: A Diffusion Model for Generating Synthetic Data in Survival Analysis.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL
SurvDiff: A Diffusion Model for Generating Synthetic Data in Survival Analysis.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL
J. Bi • D. Yan • Y. Wang • W. Huang • H. Chen • G. Wan • M. Ye • X. Xiao • H. Schütze • V. Tresp • Y. Ma
The Geometry of Reasoning: Self-Evaluation via Layerwise Trajectory Evolution.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
The Geometry of Reasoning: Self-Evaluation via Layerwise Trajectory Evolution.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
L. Ciernik • M. Morik • L. Thede • L. Eyring • S. Nakajima • Z. Akata • L. Muttenthaler
Attentive Multi-Layer Fusion for Vision Transformers.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Attentive Multi-Layer Fusion for Vision Transformers.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
W. Chen • K. Tateno • H. Matsuki • M. Niemeyer • D. Cremers • F. Tombari
DynaTok: Token-Based 4D Reconstruction from Partial Point Clouds.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
DynaTok: Token-Based 4D Reconstruction from Partial Point Clouds.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
W. Durani • P. Jahn • C. Leiber • D. B. Hoffmann • T. Seidl • C. Plant • C. Böhm
CHB: A Diagnostic Toolkit for Hardness-Aware Clustering Evaluation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
CHB: A Diagnostic Toolkit for Hardness-Aware Clustering Evaluation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
F. Eichin • Y. Du • P. Mondorf • M. Matveev • B. Plank • M. A. Hedderich
ExPLAIND: Unifying Model, Data, and Training Attribution to Study Model Behavior.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
ExPLAIND: Unifying Model, Data, and Training Attribution to Study Model Behavior.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
E. S. Eberhard • L. A. Thiede • A. Aldossary • A. Burger • N. Gao • V. Bhethanabotla • A. Aspuru-Guzik • S. Günnemann
Derivative Informed Learning of Exchange-Correlation Functionals.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Derivative Informed Learning of Exchange-Correlation Functionals.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
F. Fumagalli • L. Butler • J. S. Kang • K. Ramchandran • R. T. Witter
An Odd Estimator for Shapley Values.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. arXiv
An Odd Estimator for Shapley Values.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. arXiv
D. Frauen • A. Deviyani • M. van der Schaar • S. Feuerriegel
Nonparametric LLM Evaluation from Preference Data.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Nonparametric LLM Evaluation from Preference Data.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
L. Girrbach • Z. Akata
Sparse Autoencoders are Topic Models.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Sparse Autoencoders are Topic Models.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
L. Gosch • X. Chen • Y. Scholten • S. Günnemann
Certifying Graph Neural Networks Against Label and Structure Poisoning.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Certifying Graph Neural Networks Against Label and Structure Poisoning.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
L. Galli • C. Fox • W. Bartolomaeus • M. Schmidt • H. Rauhut
Flatland: The Adventures of Gradient Descent with Large Step Sizes.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Flatland: The Adventures of Gradient Descent with Large Step Sizes.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
X. Guo • R. Zhou • Y. Wang • Q. Zhang • C. Zhang • S. Jegelka • X. Wang • J. Chai • G. Yin • W. Lin • Y. Wang
SSL4RL: Revisiting Self-supervised Learning as Intrinsic Reward for Visual-Language Reasoning.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
SSL4RL: Revisiting Self-supervised Learning as Intrinsic Reward for Visual-Language Reasoning.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
P. Holderrieth • D. Chen • L. Eyring • I. Shah • G. Anantharaman • Y. He • Z. Akata • T. Jaakkola • N. M. Boffi • M. Simchowitz
Diamond Maps: Efficient Reward Alignment via Stochastic Flow Maps.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Diamond Maps: Efficient Reward Alignment via Stochastic Flow Maps.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
M. Hasny • L. A. Daza • K. Bressem • M. Di Folco • J. A. Schnabel
No Data? No Problem: Robust Vision-Tabular Learning with Missing Values.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
No Data? No Problem: Robust Vision-Tabular Learning with Missing Values.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Y. Huang • Z. Jiang • Y. Ma • Y. Jiang • X. Wang • Y. Zhou • Y. Hao • K. Guo • P.-Y. Chen • S. Feuerriegel • X. Zhang
ProbeLLM: Automating Principled Diagnosis of LLM Failures.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
ProbeLLM: Automating Principled Diagnosis of LLM Failures.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Y. Huang • K. Roth • Q. Bouniot • W. Xu • Z. Akata
Dissecting Multimodal In-Context Learning: Modality Asymmetries and Circuit Dynamics in modern Transformers.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL GitHub
Dissecting Multimodal In-Context Learning: Modality Asymmetries and Circuit Dynamics in modern Transformers.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL GitHub
P. Janetzky • T. Schlagenhauf • S. Feuerriegel
Continual Learning of Domain-Invariant Representations.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Continual Learning of Domain-Invariant Representations.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
F. Kapl • E. Angelis • K. Maile • J. von Oswald • S. Bauer
From Growing to Looping: A Unified View of Iterative Computation in LLMs.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
From Growing to Looping: A Unified View of Iterative Computation in LLMs.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
M. Krahn • L. Bastian • V. K. Garg • B. W. Schuller • T. Birdal
Collapsed Effective Operators for Higher-order Structures.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Collapsed Effective Operators for Higher-order Structures.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
A. Kosmala • S. Günnemann • M. Gao • B. M. Wood
Speculative Sampling For Faster Molecular Dynamics.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Speculative Sampling For Faster Molecular Dynamics.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
L. Kreitner • P. Hager • J. Mengedoht • G. Kaissis • D. Rückert • M. J. Menten
Efficient numeracy in language models through single-token number embeddings.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL
Efficient numeracy in language models through single-token number embeddings.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL
E. Kasneci • G. Kasneci
Sycophancy is an Educational Safety Risk: Why LLM Tutors Need Sycophancy Benchmarks.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Sycophancy is an Educational Safety Risk: Why LLM Tutors Need Sycophancy Benchmarks.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
G. Kaissis • D. Mildenberger • J. Gomez • M. J. Menten • E. Triantafillou
Step-resolved data attribution for looped transformers.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Step-resolved data attribution for looped transformers.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
F. Kapl • A. M. K. Mamaghan • M. Seitzer • K. H. Johansson • C. Marr • S. Bauer • A. Dittadi
Are Object-Centric Representations Better at Compositional Generalization?
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Are Object-Centric Representations Better at Compositional Generalization?
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
C. Koke • Y. Shen • A. Saroha • M. Eisenberger • B. Rieck • M. M. Bronstein • D. Cremers
Graph Neural Networks Are Not Continuous Across Graph Resolutions.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Graph Neural Networks Are Not Continuous Across Graph Resolutions.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
E. Kavak • T. N. Wolf • C. Wachinger
DISCO: Mitigating Bias in Deep Learning with Conditional Distance Correlation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL GitHub
DISCO: Mitigating Bias in Deep Learning with Conditional Distance Correlation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL GitHub
M. Li • Y. Ding • E. Garces Arias • C. Heumann
Beyond Temperature: Hyperfitting as a Late-Stage Geometric Expansion.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. arXiv
Beyond Temperature: Hyperfitting as a Late-Stage Geometric Expansion.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. arXiv
F. Lin • V. Hofmann • X. Wan • W. Wang • Z. Ding • A. G. Cohn • J. B. Pierrehumbert
Can Large Language Models Generalize Procedures Across Representations?
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Can Large Language Models Generalize Procedures Across Representations?
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
J. Li • M. Liu • J. Pan • C. Liu • W. Bai • C. I. Bercea • J. A. Schnabel
Dynamic Decision Learning: Test-Time Evolution for Abnormality Grounding in Rare Diseases.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Dynamic Decision Learning: Test-Time Evolution for Abnormality Grounding in Rare Diseases.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
V. Melnychuk • V. Balazadeh • S. Feuerriegel • R. G. Krishnan
Frequentist Consistency of Prior-Data Fitted Networks for Causal Inference.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Frequentist Consistency of Prior-Data Fitted Networks for Causal Inference.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
N. Morbitzer • J. Evers • A. Savkin • T. Stauner • N. Navab • F. Tombari • S. Gasperini
Future Dynamic 3D Reconstruction: A 3D World Model with Disentangled Ego-Motio.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Future Dynamic 3D Reconstruction: A 3D World Model with Disentangled Ego-Motio.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
M. Matveev • V. Fojtik • H.-H. Chou • G. Kutyniok • J. Maly
Conflicting Biases at the Edge of Stability: Norm versus Sharpness Regularization.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Conflicting Biases at the Edge of Stability: Norm versus Sharpness Regularization.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
H. Ma • D. Frauen • S. Feuerriegel
DeepBlip: Estimating Conditional Average Treatment Effects Over Time.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
DeepBlip: Estimating Conditional Average Treatment Effects Over Time.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
J. McGinnis • F. A. Hölzl • S. Shit • F. Bieder • P. Friedrich • M. Mühlau • B. Menze • D. Rückert • B. Wiestler
Optimizing Rank for High-Fidelity Implicit Neural Representations.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Optimizing Rank for High-Fidelity Implicit Neural Representations.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
K. Oublal • Q. Bouniot • Q. Gan • S. Clémençon • Z. Akata
TimeSAE: Sparse Decoding for Faithful Explanations of Black-Box Time Series Models.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
TimeSAE: Sparse Decoding for Faithful Explanations of Black-Box Time Series Models.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
J. Odgers • B. Riegler • S. Swaroop • V. Fortuin
Gaussian Mean Field Variational Inference can Overestimate Predictive Variance.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Gaussian Mean Field Variational Inference can Overestimate Predictive Variance.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
T. Papamarkou • P. Alquier • M. Bauer • W. Buntine • A. Davison • G. K. Dziugaite • M. Filippone • A. Y. K. Foong • V. Fortuin • D. Fouskakis • J. Frellsen • E. Hüllermeier • T. Karaletsos • M. E. Khan • N. Kotelevskii • S. Lahlou • Y. Li • F. Liu • C. Lyle • T. Möllenhoff • K. Palla • M. Panov • Y. Sale • K. Schweighofer • A. Shelmanov • S. Swaroop • M. Trapp • W. Waegeman • A. G. Wilson • A. Zaytsev
Position: Agentic AI Orchestration Should Be Bayes-Consistent.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Position: Agentic AI Orchestration Should Be Bayes-Consistent.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
M. Pach • J. Bader • Q. Bouniot • S. Belongie • Z. Akata
The Latent Color Subspace: Emergent Order in High-Dimensional Chaos.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
The Latent Color Subspace: Emergent Order in High-Dimensional Chaos.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
D. Rundel • F. Fumagalli • M. Muschalik • B. Bischl • M. Feurer
ShaplEIG: Bayesian Experimental Design for Shapley Value Estimation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. URL GitHub
ShaplEIG: Bayesian Experimental Design for Shapley Value Estimation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. URL GitHub
A. Riess • J. F. Gomez • F. Calmon • J. A. Schnabel • G. Kaissis
Optimal conversion from Rényi Differential Privacy to f-Differential Privacy.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Optimal conversion from Rényi Differential Privacy to f-Differential Privacy.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
S. Roschmann • P. Krzakala • S. Mazelet • Q. Bouniot • Z. Akata
SOTAlign: Semi-Supervised Alignment of Unimodal Vision and Language Models via Optimal Transport.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
SOTAlign: Semi-Supervised Alignment of Unimodal Vision and Language Models via Optimal Transport.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
M. Rashed • D. F. V. Madeira • B. Gholami • C. Guerbuez • Y. Yang • N. Thuerey
On the Generalization in Topology Optimization via Sensitivity-Conditioned Bernoulli Flow Matching.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
On the Generalization in Topology Optimization via Sensitivity-Conditioned Bernoulli Flow Matching.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
S. N. Ramachandran • S. Sra
Trees to Flows and Back: Unifying Decision Trees and Diffusion Models.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Trees to Flows and Back: Unifying Decision Trees and Diffusion Models.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
D. Rügamer
On the Epistemic Uncertainty of Overparametrized Neural Networks.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
On the Epistemic Uncertainty of Overparametrized Neural Networks.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
E. Sommer • K. Diao • J. Robnik • U. Seljak • D. Rügamer
Can Microcanonical Langevin Dynamics Leverage Mini-Batch Gradient Noise?
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. URL
Can Microcanonical Langevin Dynamics Leverage Mini-Batch Gradient Noise?
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. URL
E. Sommer • D. Rügamer
Position: The Time for Sampling Is Now! Charting a New Course for Bayesian Deep Learning.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL
Position: The Time for Sampling Is Now! Charting a New Course for Bayesian Deep Learning.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Spotlight Presentation. To be published. Preprint available. URL
S. Thies • V. Bengs • T. Kaufmann • S. Vollmer • E. Hüllermeier
Calibrated Preference Learning: The Case of Label Ranking.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Calibrated Preference Learning: The Case of Label Ranking.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
T. Tomov • D. Fuchsgruber • S. Günnemann
Task-Awareness Improves LLM Generations and Uncertainty.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Task-Awareness Improves LLM Generations and Uncertainty.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Z. S. Taghavi • A. Modarressi • H. Schütze • A. Marfurt
With Argus Eyes: Assessing Retrieval Gaps via Uncertainty Scoring to Detect and Remedy Retrieval Blind Spots.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
With Argus Eyes: Assessing Retrieval Gaps via Uncertainty Scoring to Detect and Remedy Retrieval Blind Spots.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
E. Valero-Leal • B. Bischl • P. Larrañaga • C. Bielza • G. Casalicchio
Optimal Transport Group Counterfactual Explanations.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Optimal Transport Group Counterfactual Explanations.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
T. D. Wang • T. Birdal • N. Navab • L. Bastian
DisPOSE: Projected Polystochastic Diffusion for Self-Supervised Multi-View 3D Human Pose Estimation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
DisPOSE: Projected Polystochastic Diffusion for Self-Supervised Multi-View 3D Human Pose Estimation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
R. T. Witter • Á. Parafita • T. Garriga • M. Muschalik • F. Fumagalli • A. Brando • L. Rosenblatt
Exactly Computing do-Shapley Values.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. arXiv
Exactly Computing do-Shapley Values.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. arXiv
J. Xiao • M. Gao • S. Weber • G. Yang • D. Cremers
Functional Attention: From Pairwise Affinities to Functional Correspondences.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Functional Attention: From Pairwise Affinities to Functional Correspondences.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
M. Zarvandi • M. Timothy • T. Wasserer • D. Ghoshdastidar
Interpretable Self-Supervised Learning via Representer Landmarks and Nyström Approximation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Interpretable Self-Supervised Learning via Representer Landmarks and Nyström Approximation.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Z. Zhang • S. Yang • B. Prenkaj • G. Kasneci
Active Tabular Augmentation via Policy-Guided Diffusion Inpainting.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Active Tabular Augmentation via Policy-Guided Diffusion Inpainting.
ICML 2026 - 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Workshops (12 papers)
V. Bürgin • D. Herbst • Y.-W. E. Lin • S. Jegelka
Beyond Structural Symmetries: Linear Mode Connectivity via Neuron Identifiability.
WSS @ICML 2026 - Workshop on Weight-Space Symmetries at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Oral Presentation. To be published. Preprint available. URL GitHub
Beyond Structural Symmetries: Linear Mode Connectivity via Neuron Identifiability.
WSS @ICML 2026 - Workshop on Weight-Space Symmetries at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. Oral Presentation. To be published. Preprint available. URL GitHub
N. Hyeon-Woo • T.-H. Oh • Z. Akata • S. Alaniz
From Numbers to Narratives: Goal-Oriented Summarization of Machine Learning Model Differences.
CompLearn @ICML 2026 - 2nd Workshop on Compositional Learning: Safety, Interpretability, and Agents at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
From Numbers to Narratives: Goal-Oriented Summarization of Machine Learning Model Differences.
CompLearn @ICML 2026 - 2nd Workshop on Compositional Learning: Safety, Interpretability, and Agents at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
S. Korkut • M. A. Bravo Sarmiento • S. Kim • Z. Akata
From Accuracy to Visual Dependence: Auditing and Filtering Modality Collapse in Traffic VideoQA.
CTB @ICML 2026 - Workshop on Combining Theory and Benchmarks: Towards A Virtuous Cycle to Understand at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
From Accuracy to Visual Dependence: Auditing and Filtering Modality Collapse in Traffic VideoQA.
CTB @ICML 2026 - Workshop on Combining Theory and Benchmarks: Towards A Virtuous Cycle to Understand at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
S. Kolli • T. Cavelius • N. Nikeghbal • S. Dalal • J. Diesner
StylisticBias: A Few Human Visual Cues Drive Most Social Bias in MLLMs.
AI4GOOD @ICML 2026 - Workshop on Trustworthy AI for Good at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
StylisticBias: A Few Human Visual Cues Drive Most Social Bias in MLLMs.
AI4GOOD @ICML 2026 - Workshop on Trustworthy AI for Good at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL GitHub
Y. Li • G. Kasneci
Safety Cost of Steering Vectors Is Separable and Reducible.
CompLearn @ICML 2026 - 2nd Workshop on Compositional Learning: Safety, Interpretability, and Agents at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Safety Cost of Steering Vectors Is Separable and Reducible.
CompLearn @ICML 2026 - 2nd Workshop on Compositional Learning: Safety, Interpretability, and Agents at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Y. Li • G. Kasneci
Safety Cost of Steering Vectors Is Separable and Reducible.
AI4GOOD @ICML 2026 - Workshop on Trustworthy AI for Good at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Safety Cost of Steering Vectors Is Separable and Reducible.
AI4GOOD @ICML 2026 - Workshop on Trustworthy AI for Good at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
N. Nikeghbal • A. H. Kargaran • S. Kolli • J. Diesner
Who Flips? Self- and Cross-Model Counterarguments Reveal Answer Instability in LLMs.
AIWILD @ICML 2026 - 2nd Workshop on Agents in the Wild: Safety, Security, and Beyond at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Who Flips? Self- and Cross-Model Counterarguments Reveal Answer Instability in LLMs.
AIWILD @ICML 2026 - 2nd Workshop on Agents in the Wild: Safety, Security, and Beyond at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
N. Nikeghbal • A. H. Kargaran • S. Kolli • J. Diesner
Who Flips? Self- and Cross-Model Counterarguments Reveal Answer Instability in LLMs.
AI4GOOD @ICML 2026 - Workshop on Trustworthy AI for Good at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Who Flips? Self- and Cross-Model Counterarguments Reveal Answer Instability in LLMs.
AI4GOOD @ICML 2026 - Workshop on Trustworthy AI for Good at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
R. Poletukhin • M. Kollovieh • E. Eberhard • S. Günnemann
3D Molecule Generation from Rigid Motifs via SE(3) Flows.
GenBio @ICML 2026 - Workshop on Generative and Agentic AI for Biology at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
3D Molecule Generation from Rigid Motifs via SE(3) Flows.
GenBio @ICML 2026 - Workshop on Generative and Agentic AI for Biology at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
V. Pauline • K. Tamogashev • A. Carter • S. Choi • S. Bauer • E. S. Whitammer
Amortised Inference through One-Step Implicit Sampling.
SPIGM @ICML 2026 - Workshop on Structured Probabilistic Inference and Generative Modeling at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. URL
Amortised Inference through One-Step Implicit Sampling.
SPIGM @ICML 2026 - Workshop on Structured Probabilistic Inference and Generative Modeling at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. URL
Y. Sale • C. Bülte • F. Czaja • J. Stiller • E. Hüllermeier
What Uncertainties Do We Need for Dynamical Systems?
EIML @ICML 2026 - Workshop on Epistemic Intelligence in Machine Learning at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. URL
What Uncertainties Do We Need for Dynamical Systems?
EIML @ICML 2026 - Workshop on Epistemic Intelligence in Machine Learning at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. URL
K. Sakalyan • S. Kim • L. Schwinn • Q. Bouniot • Z. Akata
Block-Level Recursion: Adaptive Test-Time Routing in Large Language Models.
AdaptFM @ICML 2026 - Workshop on Resource-Adaptive Foundation Model Inference at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Block-Level Recursion: Adaptive Test-Time Routing in Large Language Models.
AdaptFM @ICML 2026 - Workshop on Resource-Adaptive Foundation Model Inference at the 43rd International Conference on Machine Learning. Seoul, South Korea, Jul 06-11, 2026. To be published. Preprint available. URL
Related
02.07.2026
Study Reveals Privacy Risks in Medical AI
Study led by MCML Director Daniel Rückert reveals higher privacy risks in medical AI models.
01.07.2026
MCML at ACL 2026
MCML researchers are represented with 36 papers at ACL 2026 (20 Main, 15 Findings, and 1 Workshop).
30.06.2026
Digdeep Podcast: Live From Futuromondo With Dominik Tress: Organic Food for Kids, AI for the Rest
The new episode of #digdeep with entrepreneur Dominik Tress explores how AI is transforming processes in the organic food industry.
