Inference in High Dimensions: Light-speed Algorithms and Information Limits

Project Period: 2024-10-01 – 2029-09-30
Funder: European Research Council
Acronym
INF_2
Principal Investigator
Department(s)
Grant Number
101161364
Grant DOI
Funder
European Research Council
Funder Schema
HE-ERC-StG
Funder Registry

6 Publications

2026 | Epub ahead of print | Journal Article | IST-REx-ID: 21488 | OA
Joint modeling of whole-genome sequencing data for human height via approximate message passing
A. Depope, J. Bajzik, M. Mondelli, M.R. Robinson, Cell Genomics (2026).
[Published Version] View | DOI | Download Published Version (ext.)
 
2025 | Published | Conference Paper | IST-REx-ID: 20033 | OA
High-dimensional analysis of knowledge distillation: Weak-to-Strong generalization and scaling laws
M. Emrullah Ildiz, H.A. Gozeten, E.O. Taga, M. Mondelli, S. Oymak, in:, 13th International Conference on Learning Representations, ICLR, 2025, pp. 2967–3006.
[Published Version] View | Files available | arXiv
 
2025 | Published | Conference Paper | IST-REx-ID: 20035 | OA
Wide neural networks trained with weight decay provably exhibit neural collapse
A. Jacot, P. Súkeník, Z. Wang, M. Mondelli, in:, 13th International Conference on Learning Representations, ICLR, 2025, pp. 1905–1931.
[Published Version] View | Files available | arXiv
 
2025 | Published | Conference Paper | IST-REx-ID: 21324 | OA
Spurious correlations in high dimensional regression: The roles of regularization, simplicity bias and over-parameterization
S. Bombari, M. Mondelli, in:, Proceedings of the 42nd International Conference on Machine Learning, ML Research Press, 2025, pp. 4839–4873.
[Published Version] View | Files available | arXiv
 
2025 | Published | Conference Paper | IST-REx-ID: 21325 | OA
Test-time training provably improves transformers as in-context learners
H.A. Gozeten, M.E. Ildiz, X. Zhang, M. Soltanolkotabi, M. Mondelli, S. Oymak, in:, Proceedings of the 42nd International Conference on Machine Learning, ML Research Press, 2025, pp. 20266–20295.
[Published Version] View | Files available | PubMed | Europe PMC
 
2025 | Published | Conference Paper | IST-REx-ID: 21328 | OA
Spectral estimators for multi-index models: Precise asymptotics and optimal weak recovery
F. Kovačević, Z. Yihan, M. Mondelli, in:, Proceedings of 38th Conference on Learning Theory, ML Research Press, 2025, pp. 3354–3404.
[Published Version] View | Files available | arXiv
 

Search

Filter Publications

Display / Sort

Export / Embed