ICML 2023: Plus de 80 articles scientifiques affiliées à Mila

Photo du centre des congrès d'Hawaii

Du 23 au 29 juillet 2023, les chercheuses et chercheurs de Mila participeront à la quarantième Conférence internationale sur l'apprentissage automatique (ICML) à Hawaï. Elles et ils y partageront plus de 80 travaux de recherche lors de présentations orales, de sessions de posters et d'ateliers devant d’autres experts du monde entier.

Voici une liste d'articles présentés à ICML 2023 contenant au moins un·e auteur·rice affilié·e à Mila:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Titre

 

 

 

 

Auteur·rice·s

 

 

Cognitive Models as Simulators: Using Cognitive Models to Tap into Implicit Human FeedbackArdavan S. Nobandegani, Thomas Shultz, Irina Rish
FusionRetro: Molecule Representation Fusion via In-Context Learning for Retrosynthetic PlanningSongtao Liu, Zhengkai Tu, Minkai Xu, Zuobai Zhang, Lu Lin, Zhitao Ying, Jian Tang, Peilin Zhao, Dinghao Wu
Neural FIM for learning Fisher information metrics from point cloud dataOluwadamilola Fasina, Guillaume Huguet, Alexander Tong, Yanlei Zhang, Guy Wolf, Maximilian Nickel, Ian M. Adelstein, Smita Krishnaswamy
Mastering the Unsupervised Reinforcement Learning Benchmark from PixelsSai Rajeswar, Pietro Mazzaglia, Tim Verbelen, Alexandre Piché, Bart Dhoedt, Aaron Courville, Alexandre Lacoste
Lie Point Symmetry and Physics Informed NetworksTara Akhound-Sadegh, Laurence Perreault-Levasseur, Johannes Brandstetter, Max Welling, Siamak Ravanbakhsh
Privacy-Aware Compression for Federated Learning Through Numerical Mechanism DesignChuan Guo, Kamalika Chaudhuri, Pierre Stock, Michael Rabbat
Assessing Neural Network Representations During Training Using Data Diffusion SpectraDanqi Liao, Chen Liu, Alexander Tong, Guillaume Huguet, Guy Wolf, Maximilian Nickel, Ian M. Adelstein, Smita Krishnaswamy
Accelerating exploration and representation learning with offline pre-trainingBogdan Mazoure, Jake Bruce, Doina Precup, Rob Fergus, Ankit Anand
Sampling-Based Accuracy Testing of Posterior Estimators for General InferencePablo Lemos, A. Coogan, Yashar Hezaveh, Laurence Perreault-Levasseur
Bootstrapped Representations in Reinforcement LearningCharline Le Lan, Stephen Tu, Mark Rowland, Anna Harutyunyan, Rishabh Agarwal, Marc G. Bellemare, Will Dabney
Can Forward Gradient Match Backpropagation?Louis Fournier, Stephane Rivaud, Eugene Belilovsky, Michael Eickenberg, Edouard Oyallon
High-Probability Bounds for Stochastic Optimization and Variational Inequalities: the Case of Unbounded VarianceA. Sadiev, Marina Danilova, Eduard Gorbunov, Samuel Horv'ath, Gauthier Gidel, P. Dvurechensky, A. Gasnikov, Peter Richtarik
Equivariance With Learned Canonicalization FunctionsOumar Kaba, Arnab Kumar Mondal, Yan Zhang, Yoshua Bengio, Siamak Ravanbakhsh
Repository-Level Prompt Generation for Large Language Models of CodeDisha Shrivastava, Hugo Larochelle, Danny Tarlow
Uncertain Evidence in Probabilistic Models and Stochastic SimulatorsAndreas Munk, A. Mead, Frank Wood
Target-based Surrogates for Stochastic OptimizationJ. Wilder Lavington, Sharan Vaswani, Reza Babanezhad Harikandeh, Mark Schmidt, Nicolas Roux
Identifiability of Discretized Latent Coordinate Systems via Density Landmarks DetectionVitória Barin-Pacela, Kartik Ahuja, Simon Lacoste-Julien, Pascal Vincent
ProtST: Multi-Modality Learning of Protein Sequences and Biomedical TextsMinghao Xu, Xinyu Yuan, Santiago Miret, Jian Tang
Regions of Reliability in the Evaluation of Multivariate Probabilistic ForecastsE. Marcotte, Valentina Zantedeschi, Alexandre Drouin, Nicolas Chapados
Towards Reliable Neural SpecificationsChuqin Geng, Nham Le, Xiaojie Xu, Zhaoyue Wang, A. Gurfinkel, Xujie Si
Better Training of GFlowNets with Local Credit and Incomplete TrajectoriesL. Pan, Nikolay Malkin, Dinghuai Zhang, Yoshua Bengio
Maximal Initial Learning Rates in Deep ReLU NetworksGaurav Iyer, Boris Hanin, David Rolnick
Hidden Symmetries of ReLU NetworksJ. Grigsby, Elisenda Grigsby, Kathryn A. Lindsey, David Rolnick
Bidirectional Learning for Offline Model-based Biological Sequence DesignCan (Sam) Chen, Yingxue Zhang, Xue Liu, Mark Coates
PAC-Bayesian Generalization Bounds for Adversarial Generative ModelsSokhna Diarra Mbacke, Florence Clerc, Pascal Germain
Deep Networks as Paths on the Manifold of Neural RepresentationsRichard D. Lange, Devin Kwok, Jordan Kyle Matelsky, Xinyue Wang, David Rolnick, Konrad P. Kording
Discovering Object-Centric Generalized Value Functions From PixelsSomjit Nath, G. Subbaraj, Khimya Khetarpal, Samira E. Kahou
Convergence of Proximal Point and Extragradient-Based Methods Beyond Monotonicity: the Case of Negative ComonotonicityEduard Gorbunov, Adrien Taylor, Samuel Horv'ath, Gauthier Gidel
Mechanistic Mode ConnectivityE. S. Lubana, Eric J. Bigelow, Robert P. Dick, David Krueger, Hidenori Tanaka
Evolving Computation GraphsAndreea Deac, Jian Tang
Flexible Phase Dynamics for Bio-Plausible Contrastive LearningEzekiel Williams, C. Bredenberg, Guillaume Lajoie
Hyena Hierarchy: Towards Larger Convolutional Language ModelsMichael Poli, Stefano Massaroli, Eric Q. Nguyen, Daniel Y. Fu, Tri Dao, S. Baccus, Yoshua Bengio, S. Ermon, Christopher Re
Can We Scale Transformers to Predict Parameters of Diverse ImageNet Models?Boris Knyazev, Doha Hwang, Simon Lacoste-Julien
Nesterov Meets Optimism: Rate-Optimal Separable Minimax OptimizationChris Junchi Li, An Yuan, Gauthier Gidel, Quanquan Gu, Michael Jordan
Prototype-Sample Relation Distillation: Towards Replay-Free Continual LearningNader Asadi, Mohammad-Javad Davari, S. Mudur, Rahaf Aljundi, Eugene Belilovsky
A Group Symmetric Stochastic Differential Equation Model for Molecule Multi-modal PretrainingShengchao Liu, Weitao Du, Zhi-Ming Ma, Hongyu Guo, Jian Tang
GFlowNet-EM for learning compositional latent variable modelsEdward J. Hu, Nikolay Malkin, Moksh Jain, Katie E Everett, Alexandros Graikos, Yoshua Bengio
Interventional Causal Representation LearningKartik Ahuja, Divyat Mahajan, Yixin Wang, Yoshua Bengio
Multi-Environment Pretraining Enables Transfer to Action Limited DatasetsDavid Venuto, Sherry Yang, Pieter Abbeel, Doina Precup, Igor Mordatch, Ofir Nachum
Goal-conditioned GFlowNets for Controllable Multi-Objective Molecular DesignJulien Roy, Pierre-luc Bacon, Christopher Joseph Pal, Emmanuel Bengio
Omega: Optimistic EMA GradientsJuan Ramirez, Rohan Sukumaran, Quentin Bertrand, Gauthier Gidel
Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task LearningSebastien Lachapelle, Tristan Deleu, Divyat Mahajan, Ioannis Mitliagkas, Yoshua Bengio, Simon Lacoste-Julien, Quentin Bertrand
GFlowOut: Dropout with Generative Flow NetworksDianbo Liu, Moksh Jain, Bonaventure F. P. Dossou, Qianli Shen, Salem Lahlou, Anirudh Goyal Alias Parth Goyal, Anirudh Goyal, Nikolay Malkin, Chris Chinenye Emezue, Dinghuai Zhang, Nadhir Hassen, Xu Ji, Kenji Kawaguchi, Yoshua Bengio
FAENet: Frame Averaging Equivariant GNN for Materials ModelingAlexandre Duval, Victor Schmidt, Alex Hernandez-Garcia, Santiago Miret, Fragkiskos D. Malliaros, Yoshua Bengio, David Rolnick
Learning GFlowNets From Partial Episodes For Improved Convergence And StabilityKanika Madan, Jarrid Rector-Brooks, Maksym Korablyov, Emmanuel Bengio, Moksh Jain, A. Nica, Andrei Cristian Nica, Tom Bosc, Yoshua Bengio, Nikolay Malkin
The Statistical Benefits of Quantile Temporal-Difference Learning for Value EstimationMark Rowland, Yunhao Tang, Clare Lyle, Rémi Munos, Marc G. Bellemare, Will Dabney
Bigger, Better, Faster: Human-level Atari with human-level efficiencyMax Schwarzer, Johan Samir Obando Ceron, Aaron Courville, Marc G. Bellemare, Rishabh Agarwal, Pablo Samuel Castro
A Heat Diffusion Perspective on Geodesic Preserving Dimensionality ReductionGuillaume Huguet, Alexander Tong, Edward De Brouwer, Yanlei Zhang, Guy Wolf, Ian M. Adelstein, Smita Krishnaswamy
R-U-SURE? Uncertainty-Aware Code Suggestions By Maximizing Utility Across Random User IntentsDaniel Dun-ning Woo Johnson, Danny Tarlow, Christian J. Walder
A theory of continuous generative flow networksSalem Lahlou, Tristan Deleu, Pablo Lemos, Dinghuai Zhang, Alexandra Volokhova, Alex Hernandez-Garcia, L'ena N'ehale Ezzine, Yoshua Bengio, Nikolay Malkin
Graphically Structured Diffusion ModelsChristian Weilbach, William Harvey, Frank Wood
CrossSplit: Mitigating Label Noise Memorization through Data SplittingJihye Kim, Aristide Baratin, Yan Zhang, Simon Lacoste-Julien
Graph Inductive Biases in Transformers without Message PassingLiheng Ma, Chen Lin, Derek Lim, Adriana Romero-Soriano, P. Dokania, Mark Coates, Philip Torr, Ser-Nam Lim, S. Lim
Discrete Key-Value BottleneckFrederik Träuble, Anirudh Goyal Alias Parth Goyal, Anirudh Goyal, Nasim Rahaman, Michael C. Mozer, Kenji Kawaguchi, Yoshua Bengio, Bernhard Scholkopf
Multi-Objective GFlowNetsMoksh Jain, Sharath Chandra Raparthy, Alex Hernandez-Garcia, Jarrid Rector-Brooks, Yoshua Bengio, Santiago Miret, Emmanuel Bengio
Unlocking Slot Attention by Changing Optimal Transport CostsYan Zhang, David W Zhang, Simon Lacoste-Julien, G. Burghouts, Cees G. M. Snoek
Joint Bayesian inference of graphical structure and parameters with a single generative flow networkTristan Deleu, Mizu Nishikawa-Toomey, Jithendaraa Subramanian, Nikolay Malkin, Laurent Charlin, Yoshua Bengio
BatchGFN: Generative flow networks for batch active learningShreshth Malik, Salem Lahlou, Andrew Jesson, Moksh Jain, Nikolay Malkin, Tristan Deleu, Yoshua Bengio, Yarin Gal
Thompson sampling for improved exploration in GFlowNetsJarrid Rector-Brooks, Kanika Madan, Moksh Jain, Maksym Korablyov, Chenghao Liu, Sarath Chandar, Nikolay Malkin, Yoshua Bengio
Improving and generalizing flow-based generative models with minibatch optimal transportAlexander Tong, Nikolay Malkin, Guillaume Huguet, Yanlei Zhang, Jarrid Rector-Brooks, Kilian Fatras, Guy Wolf, Yoshua Bengio
Simulation-free Schrödinger bridges via score and flow matchingAlexander Tong, Nikolay Malkin, Kilian Fatras, Lazar Atanackovic, Yanlei Zhang, Guillaume Huguet, Guy Wolf, Yoshua Bengio
Neural Networks Are Graphs! Graph Neural Networks for Equivariant Processing of Neural NetworksDavid W Zhang, Miltiadis Kofinas, Yan Zhang, Yunlu Chen, Gertjan J. Burghouts, Cees G. M. Snoek
Maximum State Entropy Exploration using Predecessor and Successor RepresentationsArnav Kumar Jain, Lucas Lehnert, Irina Rish, Glen Berseth
SimBIG: Field-level Simulation-based Inference of Large-scale StructurePablo Lemos, Liam Parker, ChangHoon Hahn, Bruno Régaldo-Saint Blancard, Elena Massara, Shirley Ho, David Spergel, Chirag Modi, Azadeh Moradinezhad Dizgah, Michael Eickenberg, Jiamin Hou
SimBIG: Galaxy Clustering beyond the Power SpectrumChangHoon Hahn, Pablo Lemos, Bruno Régaldo-Saint Blancard, Liam Parker, Michael Eickenberg, Shirley Ho, Jiamin Hou, Elena Massara, Chirag Modi, Azadeh Moradinezhad Dizgah, David Spergel
Deep Laplacian-based Options for Temporally-Extended ExplorationMartin Klissarov and Marlos C. Machado
Time Delay Cosmography with a Neural Ratio EstimatorEve Campeau-Poirier, Laurence Perreault-Levasseur, Adam Coogan, Yashar Hezaveh
Towards Unbiased Gravitational-Wave Parameter Estimation using Score-Based Likelihood CharacterizationRonan Legin, Kaze Wong, Maximiliano Isi, Alexandre Adam, Laurence Perreault-Levasseur, Yashar Hezaveh
Diffusion Based Representation LearningSarthak Mittal, Korbinian Abstreiter, Stefan Bauer, Bernhard Scholkopf, Arash Mehrjou
Adversarial Policies Beat Superhuman Go AIsTony Tong Wang, Adam Gleave, Tom Tseng, Kellin Pelrine, Nora Belrose, Joseph Miller, Michael D Dennis, Yawen Duan, Viktor Pogrebniak, Sergey Levine, Stuart Russell
Do as your neighbors: Invariant learning through non-parametric neighbourhood matchingAndrei Liviu Nicolicioiu, Jerry Huang, Dhanya Sridhar, Aaron Courville
Learning Diverse Features in Vision Transformers for Improved GeneralizationArmand Mihai Nicolicioiu, Andrei Liviu Nicolicioiu, Bogdan Alexe, Damien Teney
Towards Out-of-Distribution Adversarial RobustnessAdam Ibrahim, Charles Guille-Escuret, Ioannis Mitliagkas, Irina Rish, David Krueger, Pouya Bashivan
Continual Pre-Training of Large Language Models: How to re-warm your model?Kshitij Gupta*, Benjamin Thérien*, Adam Ibrahim*, Mats Leon Richter, Quentin Gregory Anthony, Eugene Belilovsky, Timothée Lesort, Irina Rish
Idiolect: A Reconfigurable Voice Coding AssistantBreandan Considine, Nicholas Albion, Xujie Si
ROSA: Random Orthogonal Subspace AdaptationMarawan Gamal, Guillaume Rabusseau
GFlowNets for Causal Discovery: an OverviewCristian Dragos Manta, Edward J. Hu, Yoshua Bengio
What if We Enrich day-ahead Solar Irradiance Time Series Forecasting with Spatio-Temporal Context?Oussama Boussif, Ghait Boukachab, Dan Assouline, Stefano Massaroli, Tianle Yuan, Loubna Benabbou, Yoshua Bengio
Guiding The Last Layer in Federated Learning with Pre-Trained ModelsGwen Legate, Nicolas Bernier, Lucas Caccia, Edouard Oyallon, Eugene Belilovsky
Re-Weighted Softmax Cross-Entropy to Control Forgetting in Federated LearningGwen Legate, Lucas Caccia, Eugene Belilovsky
Learning to Optimize with Recurrent Hierarchical TransformersAbhinav Moudgil, Boris Knyazev, Guillaume Lajoie, Eugene Belilovsky
Abstracting Imperfect Information Away from Two-Player Zero-Sum GamesSamuel Sokota, Ryan D'Orazio, Chun Kai Ling, David Wu, Zico Kolter, Noam Brown