Diego Granziol

27 papers A* 2Journal 22Unranked 2

Year	Rank	Type	Title / Venue / Authors
2026	J	`jnl`	Hessian Spectral Analysis at Foundation Model Scale. CoRR Diego Granziol, Khurshid Juarev
2026	J	`jnl`	Safety-Efficacy Trade Off: Robustness against Data-Poisoning. CoRR Diego Granziol
2025	J	`jnl`	A Linear Approach to Data Poisoning. CoRR Diego Granziol, Donald Flynn
2025	J	`jnl`	Compute-Optimal LLMs Provably Generalize Better With Scale. CoRR Marc Finzi, Sanyam Kapoor, Diego Granziol, Anming Gu, Christopher De Sa, J. Zico Kolter, Andrew Gordon Wilson
2025	A*	`conf`	Compute-Optimal LLMs Provably Generalize Better with Scale. ICLR Marc Anton Finzi, Sanyam Kapoor, Diego Granziol, Anming Gu, Christopher De Sa, J. Zico Kolter, Andrew Gordon Wilson
2025	J	`jnl`	HessFormer: Hessians at Foundation Scale. CoRR Diego Granziol
2024	J	`jnl`	Iterate Averaging in the Quest for Best Test Error. J. Mach. Learn. Res. Diego Granziol, Nicholas P. Baskerville, Xingchen Wan, Samuel Albanie, Stephen Roberts
2022	J	`jnl`	Learning Rates as a Function of Batch Size: A Random Matrix Theory Approach to Neural Network Training. J. Mach. Learn. Res. Diego Granziol, Stefan Zohren, Stephen Roberts
2022	J	`jnl`	Maximum Entropy Approach to Massive Graph Spectrum Learning with Applications. Algorithms Diego Granziol, Binxin Ru, Xiaowen Dong, Stefan Zohren, Michael A. Osborne, Stephen J. Roberts
2022	J	`jnl`	Universal characteristics of deep neural network loss surfaces from random matrix theory. CoRR Nicholas P. Baskerville, Jonathan P. Keating, Francesco Mezzadri, Joseph Najnudel, Diego Granziol
2021	J	`jnl`	Applicability of Random Matrix Theory in Deep Learning. CoRR Nicholas P. Baskerville, Diego Granziol, Jonathan P. Keating
2020	J	`jnl`	Beyond Random Matrix Theory for Deep Networks. CoRR Diego Granziol
2020	J	`jnl`	Curvature is Key: Sub-Sampled Loss Surfaces and the Implications for Large Batch Training. CoRR Diego Granziol
2020	J	`jnl`	Explaining the Adaptive Generalisation Gap. CoRR Diego Granziol, Samuel Albanie, Xingchen Wan, Stephen J. Roberts
2020	J	`jnl`	Flatness is a False Friend. CoRR Diego Granziol
2020	J	`jnl`	Iterate Averaging Helps: An Alternative Perspective in Deep Learning. CoRR Diego Granziol, Xingchen Wan, Stephen Roberts
2020	—	`—`	Spectral machine learning. Diego Granziol
2019	J	`jnl`	A Maximum Entropy approach to Massive Graph Spectra. CoRR Diego Granziol, Robin Ru, Stefan Zohren, Xiaowen Dong, Michael A. Osborne, Stephen J. Roberts
2019	J	`jnl`	MEMe: An Accurate Maximum Entropy Method for Efficient Approximations in Large-Scale Machine Learning. CoRR Diego Granziol, Bin Xin Ru, Stefan Zohren, Xiaowen Dong, Michael A. Osborne, Stephen J. Roberts
2019	J	`jnl`	MEMe: An Accurate Maximum Entropy Method for Efficient Approximations in Large-Scale Machine Learning. Entropy Diego Granziol, Bin Xin Ru, Stefan Zohren, Xiaowen Dong, Michael A. Osborne, Stephen J. Roberts
2019	J	`jnl`	MLRG Deep Curvature. CoRR Diego Granziol, Xingchen Wan, Timur Garipov, Dmitry P. Vetrov, Stephen Roberts
2018	J	`jnl`	Entropic Spectral Learning in Large Scale Networks. CoRR Diego Granziol, Bin Xin Ru, Stefan Zohren, Xiaowen Dong, Michael A. Osborne, Stephen J. Roberts
2018	A*	`conf`	Fast Information-theoretic Bayesian Optimisation. ICML Bin Xin Ru, Mark McLeod, Diego Granziol, Michael A. Osborne
2018	J	`jnl`	VBALD - Variational Bayesian Approximation of Log Determinants. CoRR Diego Granziol, Edward Wagstaff, Bin Xin Ru, Michael A. Osborne, Stephen J. Roberts
2017	—	`conf`	Entropic Trace Estimates for Log Determinants. ECML/PKDD (1) Jack K. Fitzsimons, Diego Granziol, Kurt Cutajar, Michael A. Osborne, Maurizio Filippone, Stephen J. Roberts
2017	J	`jnl`	Entropic Trace Estimates for Log Determinants. CoRR Jack K. Fitzsimons, Diego Granziol, Kurt Cutajar, Michael A. Osborne, Maurizio Filippone, Stephen J. Roberts
2017	—	`conf`	Entropic determinants of massive matrices. IEEE BigData Diego Granziol, Stephen J. Roberts