DOI,IST REx ID,Research Group,Title of publication
null,18890,"GradSch,MaMo",Average gradient outer product as a mechanism for deep neural collapse
null,18891,"GradSch,MaMo,ChLa",Neural collapse versus low-rank bias: Is deep neural collapse really optimal?
null,18897,"JaMa,MaMo",Improved convergence of score-based diffusion models via prediction-correction
null,18972,MaMo,How spurious features are memorized: Precise analysis for random and NTK features
null,18973,MaMo,Towards understanding the word sensitivity of attention layers: A study via random features
null,19518,"DaAl,MaMo",The iterative optimal brain surgeon: Faster sparse recovery by leveraging second-order information
10.1109/TIT.2024.3367767,15172,MaMo,Concentration without independence via information measures
10.1109/ICASSP48485.2024.10447198,17147,"MaMo,MaRo",Inference of genetic effects via approximate message passing
10.1109/TIT.2024.3430842,17330,MaMo,Zero-rate thresholds and new capacity bounds for list-decoding and list-recovery
10.48550/arXiv.2305.14164,17350,"JaMa,MaMo",Improved convergence of score-based diffusion models via prediction-correction
10.15479/at:ista:17465,17465,"GradSch,DaAl,MaMo",High-dimensional limits in artificial neural networks
null,17469,"DaAl,MaMo",Compression of structured data with autoencoders: Provable benefit of nonlinearities and depth
10.4230/LIPIcs.ICALP.2023.99,14083,MaMo,Zero-rate thresholds and new capacity bounds for list-decoding and list-recovery
10.1109/TIT.2023.3260950,12838,MaMo,Multiple packing: Lower bounds via infinite constellations
null,12859,"GradSch,MaMo",Beyond the universal law of robustness: Sharper laws for random features and neural tangent kernels
10.1109/TIT.2023.3292219,13269,MaMo,Codes for the Z-channel
10.1073/pnas.2302028120,13315,MaMo,Fundamental limits in structured principal component analysis and how to reach them
10.1109/ITW55543.2023.10160238,13321,MaMo,Approximate message passing for multi-layer estimation in rotationally invariant models
10.1109/tit.2023.3257239,14751,MaMo,Zero-error communication over adversarial MACs
null,14921,"MaMo,ChLa",Deep neural collapse is provably optimal for the deep unconstrained features model