Carlo Luschi

35 papers A* 6Misc 1Journal 28
YearRankTypeTitle / Venue / Authors
2026 J jnl
MXNorm: Reusing MXFP block scales for efficient tensor normalisation.
CoRR
Callum McLean, Luke Y. Prince, Alexandre Payot, Paul Balanca, Carlo Luschi
2026 J jnl
UltRAG: a Universal Simple Scalable Recipe for Knowledge Graph RAG.
CoRR
Dobrik Georgiev, Kheeran Naidu, Alberto Cattaneo, Federico Monti, Carlo Luschi, Daniel Justus
2025 J jnl
Elucidating the Design Space of FP4 training.
CoRR
Robert Hu, Carlo Luschi, Paul Balanca
2025 J jnl
Ground-Truth Subgraphs for Better Training and Evaluation of Knowledge Graph Augmented LLMs.
CoRR
Alberto Cattaneo, Carlo Luschi, Daniel Justus
2025 J jnl
Optimal Formats for Weight Quantisation.
CoRR
Douglas Orr, Luka Ribar, Carlo Luschi
2025 J jnl
The role of graph topology in the performance of biomedical Knowledge Graph Completion models.
Bioinform.
Alberto Cattaneo, Stephen Bonner, Thomas Martynec, Edward R. Morrissey, Carlo Luschi, Ian P. Barrett, Daniel Justus
2025 A* conf
u-μP: The Unit-Scaled Maximal Update Parametrization.
ICLR
Charlie Blake, Constantin Eichenberg, Josef Dean, Lukas Balles, Luke Yuri Prince, Björn Deiseroth, Andrés Felipe Cruz-Salinas, Carlo Luschi, Samuel Weinbach, Douglas Orr
2024 J jnl
Reducing the Cost of Quantum Chemical Data By Backpropagating Through Density Functional Theory.
CoRR
Alexander Mathiasen, Hatem Helal, Paul Balanca, Adam Krzywaniak, Ali Parviz, Frederik Hvilshøj, Blazej Banaszewski, Carlo Luschi, Andrew William Fitzgibbon
2024 J jnl
Scalify: scale propagation for efficient low-precision LLM training.
CoRR
Paul Balanca, Sam Hosegood, Carlo Luschi, Andrew W. Fitzgibbon
2024 A* conf
SparQ Attention: Bandwidth-Efficient LLM Inference.
ICML
Luka Ribar, Ivan Chelombiev, Luke Hudlass-Galley, Charlie Blake, Carlo Luschi, Douglas Orr
2024 J jnl
The Role of Graph Topology in the Performance of Biomedical Knowledge Graph Completion Models.
CoRR
Alberto Cattaneo, Stephen Bonner, Thomas Martynec, Carlo Luschi, Ian P. Barrett, Daniel Justus
2024 J jnl
u-μP: The Unit-Scaled Maximal Update Parametrization.
CoRR
Charlie Blake, Constantin Eichenberg, Josef Dean, Lukas Balles, Luke Y. Prince, Björn Deiseroth, Andrés Felipe Cruz-Salinas, Carlo Luschi, Samuel Weinbach, Douglas Orr
2023 A* conf
Generating QM1B with PySCF
NeurIPS
Alexander Mathiasen, Hatem Helal, Kerstin Klaser, Paul Balanca, Josef Dean, Carlo Luschi, Dominique Beaini, Andrew W. Fitzgibbon, Dominic Masters
2023 J jnl
Generating QM1B with PySCF
CoRR
Alexander Mathiasen, Hatem Helal, Kerstin Klaser, Paul Balanca, Josef Dean, Carlo Luschi, Dominique Beaini, Andrew W. Fitzgibbon, Dominic Masters
2023 J jnl
SparQ Attention: Bandwidth-Efficient LLM Inference.
CoRR
Luka Ribar, Ivan Chelombiev, Luke Hudlass-Galley, Charlie Blake, Carlo Luschi, Douglas Orr
2023 J jnl
Training and inference of large language models using 8-bit floating point.
CoRR
Sergio P. Perez, Yan Zhang, James Briggs, Charlie Blake, Josh Levy-Kramer, Paul Balanca, Carlo Luschi, Stephen Barlow, Andrew William Fitzgibbon
2023 A* conf
Unit Scaling: Out-of-the-Box Low-Precision Training.
ICML
Charlie Blake, Douglas Orr, Carlo Luschi
2023 J jnl
Unit Scaling: Out-of-the-Box Low-Precision Training.
CoRR
Charlie Blake, Douglas Orr, Carlo Luschi
2022 J jnl
8-bit Numerical Formats for Deep Neural Networks.
CoRR
Badreddine Noune, Philip Jones, Daniel Justus, Dominic Masters, Carlo Luschi
2022 J jnl
BESS: Balanced Entity Sampling and Sharing for Large-Scale Knowledge Graph Completion.
CoRR
Alberto Cattaneo, Daniel Justus, Harry Mellor, Douglas Orr, Jérôme Maloberti, Zhenying Liu, Thorin Farnsworth, Andrew W. Fitzgibbon, Blazej Banaszewski, Carlo Luschi
2021 J jnl
GroupBERT: Enhanced Transformer Architecture with Efficient Grouped Structures.
CoRR
Ivan Chelombiev, Daniel Justus, Douglas Orr, Anastasia Dietrich, Frithjof Gressmann, Alexandros Koliousis, Carlo Luschi
2021 J jnl
Making EfficientNet More Efficient: Exploring Batch-Independent Normalization, Group Convolutions and Reduced Resolution Training.
CoRR
Dominic Masters, Antoine Labatie, Zach Eaton-Rosen, Carlo Luschi
2021 A* conf
Proxy-Normalizing Activations to Match Batch Normalization while Removing Batch Dependence.
NeurIPS
Antoine Labatie, Dominic Masters, Zach Eaton-Rosen, Carlo Luschi
2021 J jnl
Proxy-Normalizing Activations to Match Batch Normalization while Removing Batch Dependence.
CoRR
Antoine Labatie, Dominic Masters, Zach Eaton-Rosen, Carlo Luschi
2021 J jnl
Towards Structured Dynamic Sparse Pre-Training of BERT.
CoRR
Anastasia Dietrich, Frithjof Gressmann, Douglas Orr, Ivan Chelombiev, Daniel Justus, Carlo Luschi
2020 A* conf
Improving Neural Network Training in Low Dimensional Random Bases.
NeurIPS
Frithjof Gressmann, Zach Eaton-Rosen, Carlo Luschi
2020 J jnl
Improving Neural Network Training in Low Dimensional Random Bases.
CoRR
Frithjof Gressmann, Zach Eaton-Rosen, Carlo Luschi
2020 J jnl
Parallel Training of Deep Networks with Local Updates.
CoRR
Michael Laskin, Luke Metz, Seth Nabarrao, Mark Saroufim, Badreddine Noune, Carlo Luschi, Jascha Sohl-Dickstein, Pieter Abbeel
2018 J jnl
Revisiting Small Batch Training for Deep Neural Networks.
CoRR
Dominic Masters, Carlo Luschi
2005 J jnl
Estimation of the output error statistics of space-time equalization in an antenna array EGPRS receiver with soft-decision decoding.
IEEE Trans. Wirel. Commun.
Carlo Luschi, Bernard Mulgrew
2003 J jnl
Nonparametric trellis equalization in the presence of non-Gaussian interference.
IEEE Trans. Commun.
Carlo Luschi, Bernard Mulgrew
2000 J jnl
Advanced signal-processing algorithms for energy-efficient wireless communications.
Proc. IEEE
Carlo Luschi, Magnus Sandell, Paul Strauch, Jian-Jun Wu, Costel Ilas, Ping-Wen Ong, Romain Baeriswyl, Frederic Battaglia, Spyros Karageorgis, Ran-Hong Yan
1999 Misc conf
Low complexity source controlled channel decoding in a GSM system.
ICASSP
Paul Strauch, Carlo Luschi, Magnus Sandell, Ran Yan
1996 J jnl
Joint clock recovery and baseband combining for the diversity radio channel.
IEEE Trans. Commun.
Franco Guglielmi, Carlo Luschi, Arnaldo Spalvieri
1995 J jnl
Joint Clock Recovery and Baseband Combining for the Diversity Radio Channel.
IEEE Trans. Commun.
Franco Guglielmi, Carlo Luschi, Arnaldo Spalvieri