Managing extreme AI risks amid rapid progress
Comments Published in Science: https://www.science.org/doi/10.1126/science.adn0117
作者
Yoshua Bengio
Machine Learning
Metacognitive Capabilities of LLMs: An Exploration in Mathematical Problem Solving
Comments Preprint. Under review
Generative Active Learning for the Search of Small-molecule Protein Binders
Towards DNA-Encoded Library Generation with GFlowNets
PhyloGFN: Phylogenetic inference with generative flow networks
Local Search GFlowNets
Comments ICLR 2024 (Spotlight paper), 18 pages, 17 figures
Language Models Can Reduce Asymmetry in Information Markets
Consciousness-Inspired Spatio-Temporal Abstractions for Better Generalization in Reinforcement Learning
Comments ICLR 2024 Camera-Ready
Amortizing intractable inference in large language models
Comments ICLR 2024; 23 pages; code: https://github.com/GFNOrg/gfn-lm-tuning
Expected flow networks in stochastic environments and two-player zero-sum games
Comments ICLR 2024; code: https://github.com/GFNOrg/AdversarialFlowNetworks
Delta-AI: Local objectives for amortized inference in sparse graphical models
Comments ICLR 2024; 19 pages, code: https://github.com/GFNOrg/Delta-AI/
A Hitchhiker's Guide to Geometric GNNs for 3D Atomic Systems
Simulation-free Schrödinger bridges via score and flow matching
Comments AISTATS 2024. Code: https://github.com/atong01/conditional-flow-matching
Improving and generalizing flow-based generative models with minibatch optimal transport
Comments TMLR. Code: https://github.com/atong01/conditional-flow-matching
PhAST: Physics-Aware, Scalable, and Task-specific GNNs for Accelerated Catalyst Design
Comments Journal of Machine Learning Research (JMLR)
Diffusion Generative Flow Samplers: Improving learning signals through partial trajectory optimization
Comments Accepted by ICLR 2024
Machine learning and information theory concepts towards an AI Mathematician
Comments To appear in the Bulletin of the AMS, 2024
Tree Cross Attention
Comments Accepted by ICLR 2024
Interventional Causal Representation Learning
Distributional GFlowNets with Quantile Flows
Comments Accepted by TMLR
Computing Power and the Governance of Artificial Intelligence
Comments Figures can be accessed at: https://github.com/lheim/CPGAI-Figures
GEO-Bench: Toward Foundation Models for Earth Monitoring
Comments arXiv admin note: text overlap with arXiv:2112.00570
DynGFN: Towards Bayesian Inference of Gene Regulatory Networks with GFlowNets
Cycle Consistency Driven Object Discovery
Improving Gradient-guided Nested Sampling for Posterior Inference
Comments 10 pages, 5 figures. Code available at https://github.com/Pablo-Lemos/GGNS
InfoBot: Transfer and Exploration via the Information Bottleneck
Comments Accepted at ICLR'19
Let the Flows Tell: Solving Graph Combinatorial Optimization Problems with GFlowNets
Comments Accepted by NeurIPS 2023 as spotlight
Leveraging Diffusion Disentangled Representations to Mitigate Shortcuts in Underspecified Visual Tasks
Comments Accepted at Neural Information Processing Systems(NeurIPS) 2023 - Workshop on Diffusion Models
HyenaDNA: Long-Range Genomic Sequence Modeling at Single Nucleotide Resolution
Comments NeurIPS 2023 (Spotlight)
