作者
Yoshua Bengio
Machine Learning
Constant Memory Attention Block
Comments Workshop version of arXiv:2305.14567
Sources of Richness and Ineffability for Phenomenally Conscious States
Better Training of GFlowNets with Local Credit and Incomplete Trajectories
Comments ICML 2023
Discrete Key-Value Bottleneck
Comments 40th International Conference on Machine Learning (ICML 2023)
Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task Learning
Comments Appears in: Fortieth International Conference on Machine Learning (ICML 2023). 36 pages
Spotlight Attention: Robust Object-Centric Learning With a Spatial Locality Prior
Comments 16 pages, 3 figures, under review at NeurIPS 2023
GFlowNet-EM for learning compositional latent variable models
Comments ICML 2023; code: https://github.com/GFNOrg/GFlowNet-EM
Learning GFlowNets from partial episodes for improved convergence and stability
Comments ICML 2023
A theory of continuous generative flow networks
Comments ICML 2023; 32 pages; code: https://github.com/saleml/continuous-gfn
Biological Sequence Design with GFlowNets
Comments ICML 2022. 15 pages, 3 figures. Code available at: https://github.com/MJ10/BioSeq-GFN-AL. Updated GFP results
FAENet: Frame Averaging Equivariant GNN for Materials Modeling
Comments Accepted at ICML 2023
Hyena Hierarchy: Towards Larger Convolutional Language Models
Comments Additional details
Predictive Inference with Feature Conformal Prediction
Comments Published as a conference paper at ICLR 2023
RECOVER: sequential model optimization platform for combination drug repurposing identifies novel synergistic compounds in vitro
GFlowNets and variational inference
Comments ICLR 2023 final version; code: https://github.com/GFNOrg/GFN_vs_HVI
Latent Bottlenecked Attentive Neural Processes
Toward Next-Generation Artificial Intelligence: Catalyzing the NeuroAI Revolution
Comments White paper, 10 pages + 8 pages of references, 1 figures
Latent State Marginalization as a Low-cost Approach for Improving Exploration
Comments Accepted by ICLR 2023
Boosting Exploration in Multi-Task Reinforcement Learning using Adversarial Networks
DEUP: Direct Epistemic Uncertainty Prediction
Unifying Generative Models with GFlowNets and Beyond
Comments expanded version of the ICML 2022 workshop paper
Leveraging the Third Dimension in Contrastive Learning
Regeneration Learning: A Learning Paradigm for Data Generation
Structured Sparsity Inducing Adaptive Optimizers for Deep Learning
Lookback for Learning to Branch
Comments Published in Transactions on Machine Learning Research (TMLR)
Benchmarking Graph Neural Networks
Comments Benchmarking framework on GitHub at https://github.com/graphdeeplearning/benchmarking-gnns
Journal ref Journal of Machine Learning Research (JMLR), 2022
Posterior samples of source galaxies in strong gravitational lenses with score-based priors
Comments 5+6 pages, 3 figures, Accepted (poster + contributed talk) for the Machine Learning and the Physical Sciences Workshop at the 36th conference on Neural Information Processing Systems (NeurIPS 2022); Corrected style file and added authors checklist
Controlled Sparsity via Constrained Optimization or: How I Learned to Stop Tuning Penalties and Love Constraints
Comments NeurIPS 2022 - Code available at https://github.com/gallego-posada/constrained_sparsity
The Effect of Diversity in Meta-Learning
Comments Accepted at AAAI 23
