Andrew M. Saxe

89 papers A* 26A 1B 6Journal 50Unranked 6

Year	Rank	Type	Title / Venue / Authors
2026	J	`jnl`	A Theory of How Pretraining Shapes Inductive Bias in Fine-Tuning. CoRR Nicolas Anguita, Francesco Locatello, Andrew M. Saxe, Marco Mondelli, Flavia Mancini, Samuel Lippl, Clémentine C. J. Dominé
2026	J	`jnl`	Optimal Learning Rate Schedule for Balancing Effort and Performance. CoRR Valentina Njaradi, Rodrigo Carrasco-Davis, Peter E. Latham, Andrew M. Saxe
2025	A*	`conf`	A Theory of Initialisation's Impact on Specialisation. ICLR Devon Jarvis, Sebastian Lee, Clémentine Carla Juliette Dominé, Andrew M. Saxe, Stefano Sarao Mannelli
2025	J	`jnl`	A Theory of Initialisation's Impact on Specialisation. CoRR Devon Jarvis, Sebastian Lee, Clémentine C. J. Dominé, Andrew M. Saxe, Stefano Sarao Mannelli
2025	A*	`conf`	Algorithm Development in Neural Networks: Insights from the Streaming Parity Task. ICML Loek van Rossem, Andrew M. Saxe
2025	J	`jnl`	Algorithm Development in Neural Networks: Insights from the Streaming Parity Task. CoRR Loek van Rossem, Andrew M. Saxe
2025	J	`jnl`	Distinct Computations Emerge From Compositional Curricula in In-Context Learning. CoRR Jin Hwa Lee, Andrew K. Lampinen, Aaditya K. Singh, Andrew M. Saxe
2025	A*	`conf`	From Lazy to Rich: Exact Learning Dynamics in Deep Linear Networks. ICLR Clémentine Carla Juliette Dominé, Nicolas Anguita, Alexandra Maria Proca, Lukas Braun, Daniel Kunin, Pedro A. M. Mediano, Andrew M. Saxe
2025	A*	`conf`	Make Haste Slowly: A Theory of Emergent Structured Mixed Selectivity in Feature Learning ReLU Networks. ICLR Devon Jarvis, Richard Klein, Benjamin Rosman, Andrew M. Saxe
2025	J	`jnl`	Make Haste Slowly: A Theory of Emergent Structured Mixed Selectivity in Feature Learning ReLU Networks. CoRR Devon Jarvis, Richard Klein, Benjamin Rosman, Andrew M. Saxe
2025	J	`jnl`	Nonlinear dynamics of localization in neural receptive fields. CoRR Leon Lufkin, Andrew M. Saxe, Erin Grant
2025	A*	`conf`	Not all solutions are created equal: An analytical dissociation of functional and representational similarity in deep linear neural networks. ICML Lukas Braun, Erin Grant, Andrew M. Saxe
2025	J	`jnl`	Revisiting the Role of Relearning in Semantic Dementia. CoRR Devon Jarvis, Verena Klar, Richard Klein, Benjamin Rosman, Andrew M. Saxe
2025	J	`jnl`	Saddle-to-Saddle Dynamics Explains A Simplicity Bias Across Neural Network Architectures. CoRR Yedi Zhang, Andrew M. Saxe, Peter E. Latham
2025	J	`jnl`	Softmax ≥ Linear: Transformers may learn to classify in-context by kernel gradient descent. CoRR Sara Dragutinovic, Andrew M. Saxe, Aaditya K. Singh
2025	A*	`conf`	Strategy Coopetition Explains the Emergence and Transience of In-Context Learning. ICML Aaditya K. Singh, Ted Moskovitz, Sara Dragutinovic, Felix Hill, Stephanie C. Y. Chan, Andrew M. Saxe
2025	J	`jnl`	Strategy Coopetition Explains the Emergence and Transience of In-Context Learning. CoRR Aaditya K. Singh, Ted Moskovitz, Sara Dragutinovic, Felix Hill, Stephanie C. Y. Chan, Andrew M. Saxe
2025	A*	`conf`	Training Dynamics of In-Context Learning in Linear Attention. ICML Yedi Zhang, Aaditya K. Singh, Peter E. Latham, Andrew M. Saxe
2025	J	`jnl`	Training Dynamics of In-Context Learning in Linear Attention. CoRR Yedi Zhang, Aaditya K. Singh, Peter E. Latham, Andrew M. Saxe
2025	J	`jnl`	When Are Bias-Free ReLU Networks Effectively Linear Networks? Trans. Mach. Learn. Res. Yedi Zhang, Andrew M. Saxe, Peter E. Latham
2024	J	`jnl`	Abrupt and spontaneous strategy switches emerge in simple regularised neural networks. PLoS Comput. Biol. Anika T. Löwe, Léo Touzo, Paul S. Muhle-Karbe, Andrew M. Saxe, Christopher Summerfield, Nicolas W. Schuck
2024	J	`jnl`	Early learning of the optimal constant solution in neural networks and humans. CoRR Jirko Rubruck, Jan P. Bauer, Andrew M. Saxe, Christopher Summerfield
2024	A*	`conf`	Flexible task abstractions emerge in linear networks with fast and bounded units. NeurIPS Kai Sandbrink, Jan P. Bauer, Alexandra Maria Proca, Andrew M. Saxe, Christopher Summerfield, Ali Hummos
2024	J	`jnl`	Flexible task abstractions emerge in linear networks with fast and bounded units. CoRR Kai Sandbrink, Jan P. Bauer, Alexandra Maria Proca, Andrew M. Saxe, Christopher Summerfield, Ali Hummos
2024	J	`jnl`	From Lazy to Rich: Exact Learning Dynamics in Deep Linear Networks. CoRR Clémentine C. J. Dominé, Nicolas Anguita, Alexandra Maria Proca, Lukas Braun, Daniel Kunin, Pedro A. M. Mediano, Andrew M. Saxe
2024	A*	`conf`	Get rich quick: exact solutions reveal how unbalanced initializations promote rapid feature learning. NeurIPS Daniel Kunin, Allan Raventós, Clémentine C. J. Dominé, Feng Chen, David A. Klindt, Andrew M. Saxe, Surya Ganguli
2024	J	`jnl`	Get rich quick: exact solutions reveal how unbalanced initializations promote rapid feature learning. CoRR Daniel Kunin, Allan Raventós, Clémentine C. J. Dominé, Feng Chen, David A. Klindt, Andrew M. Saxe, Surya Ganguli
2024	A*	`conf`	Nonlinear dynamics of localization in neural receptive fields. NeurIPS Leon Lufkin, Andrew M. Saxe, Erin Grant
2024	J	`jnl`	On The Specialization of Neural Modules. CoRR Devon Jarvis, Richard Klein, Benjamin Rosman, Andrew M. Saxe
2024	A*	`conf`	Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks. ICML Stefano Sarao Mannelli, Yaraslau Ivashinka, Andrew M. Saxe, Luca Saglietti
2024	J	`jnl`	Tilting the Odds at the Lottery: the Interplay of Overparameterisation and Curricula in Neural Networks. CoRR Stefano Sarao Mannelli, Yaraslau Ivashinka, Andrew M. Saxe, Luca Saglietti
2024	A*	`conf`	Understanding Unimodal Bias in Multimodal Deep Linear Networks. ICML Yedi Zhang, Peter E. Latham, Andrew M. Saxe
2024	A*	`conf`	What needs to go right for an induction head? A mechanistic study of in-context learning circuits and their formation. ICML Aaditya K. Singh, Ted Moskovitz, Felix Hill, Stephanie C. Y. Chan, Andrew M. Saxe
2024	J	`jnl`	What needs to go right for an induction head? A mechanistic study of in-context learning circuits and their formation. CoRR Aaditya K. Singh, Ted Moskovitz, Felix Hill, Stephanie C. Y. Chan, Andrew M. Saxe
2024	J	`jnl`	When Are Bias-Free ReLU Networks Like Linear Networks? CoRR Yedi Zhang, Andrew M. Saxe, Peter E. Latham
2024	A*	`conf`	When Representations Align: Universality in Representation Learning Dynamics. ICML Loek van Rossem, Andrew M. Saxe
2024	J	`jnl`	When Representations Align: Universality in Representation Learning Dynamics. CoRR Loek van Rossem, Andrew M. Saxe
2024	A*	`conf`	Why Do Animals Need Shaping? A Theory of Task Composition and Curriculum Learning. ICML Jin Hwa Lee, Stefano Sarao Mannelli, Andrew M. Saxe
2023	J	`jnl`	A Theory of Unimodal Bias in Multimodal Learning. CoRR Yedi Zhang, Peter E. Latham, Andrew M. Saxe
2023	A	`conf`	Know your audience: specializing grounded language models with listener subtraction. EACL Aaditya K. Singh, David Ding, Andrew M. Saxe, Felix Hill, Andrew K. Lampinen
2023	J	`jnl`	Meta-Learning Strategies through Value Maximization in Neural Networks. CoRR Rodrigo Carrasco-Davis, Javier Alejandro Masís, Andrew M. Saxe
2023	J	`jnl`	Modelling continual learning in humans with Hebbian context gating and exponentially decaying task signals. PLoS Comput. Biol. Timo Flesch, David G. Nagy, Andrew M. Saxe, Christopher Summerfield
2023	A*	`conf`	On The Specialization of Neural Modules. ICLR Devon Jarvis, Richard Klein, Benjamin Rosman, Andrew M. Saxe
2023	J	`jnl`	Regularised neural networks mimic human insight. CoRR Anika T. Löwe, Léo Touzo, Paul S. Muhle-Karbe, Andrew M. Saxe, Christopher Summerfield, Nicolas W. Schuck
2023	J	`jnl`	The RL Perceptron: Generalisation Dynamics of Policy Learning in High Dimensions. CoRR Nishil Patel, Sebastian Lee, Stefano Sarao Mannelli, Sebastian Goldt, Andrew M. Saxe
2023	A*	`conf`	The Transient Nature of Emergent In-Context Learning in Transformers. NeurIPS Aaditya K. Singh, Stephanie C. Y. Chan, Ted Moskovitz, Erin Grant, Andrew M. Saxe, Felix Hill
2023	J	`jnl`	The Transient Nature of Emergent In-Context Learning in Transformers. CoRR Aaditya K. Singh, Stephanie C. Y. Chan, Ted Moskovitz, Erin Grant, Andrew M. Saxe, Felix Hill
2022	A*	`conf`	An Analytical Theory of Curriculum Learning in Teacher-Student Networks. NeurIPS Luca Saglietti, Stefano Sarao Mannelli, Andrew M. Saxe
2022	J	`jnl`	Continual task learning in natural and artificial agents. CoRR Timo Flesch, Andrew M. Saxe, Christopher Summerfield
2022	A*	`conf`	Exact learning dynamics of deep linear networks with prior knowledge. NeurIPS Lukas Braun, Clémentine C. J. Dominé, James FitzGerald, Andrew M. Saxe
2022	J	`jnl`	Know your audience: specializing grounded language models with the game of Dixit. CoRR Aaditya K. Singh, David Ding, Andrew M. Saxe, Felix Hill, Andrew K. Lampinen
2022	J	`jnl`	Maslow's Hammer for Catastrophic Forgetting: Node Re-Use vs Node Activation. CoRR Sebastian Lee, Stefano Sarao Mannelli, Claudia Clopath, Sebastian Goldt, Andrew M. Saxe
2022	A*	`conf`	Maslow's Hammer in Catastrophic Forgetting: Node Re-Use vs. Node Activation. ICML Sebastian Lee, Stefano Sarao Mannelli, Claudia Clopath, Sebastian Goldt, Andrew M. Saxe
2022	J	`jnl`	Modelling continual learning in humans with Hebbian context gating and exponentially decaying task signals. CoRR Timo Flesch, David G. Nagy, Andrew M. Saxe, Christopher Summerfield
2022	J	`jnl`	Probing transfer learning with a model of synthetic correlated datasets. Mach. Learn. Sci. Technol. Federica Gerace, Luca Saglietti, Stefano Sarao Mannelli, Andrew M. Saxe, Lenka Zdeborová
2022	A*	`conf`	The Neural Race Reduction: Dynamics of Abstraction in Gated Networks. ICML Andrew M. Saxe, Shagun Sodhani, Sam Jay Lewallen
2022	J	`jnl`	The Neural Race Reduction: Dynamics of Abstraction in Gated Networks. CoRR Andrew M. Saxe, Shagun Sodhani, Sam Jay Lewallen
2021	J	`jnl`	An Analytical Theory of Curriculum Learning in Teacher-Student Networks. CoRR Luca Saglietti, Stefano Sarao Mannelli, Andrew M. Saxe
2021	A*	`conf`	Continual Learning in the Teacher-Student Setup: Impact of Task Similarity. ICML Sebastian Lee, Sebastian Goldt, Andrew M. Saxe
2021	J	`jnl`	Continual Learning in the Teacher-Student Setup: Impact of Task Similarity. CoRR Sebastian Lee, Sebastian Goldt, Andrew M. Saxe
2021	B	`conf`	Inferring Actions, Intentions, and Causal Relations in a Deep Neural Network. CogSci Keno Juechems, Andrew M. Saxe
2021	J	`jnl`	Probing transfer learning with a model of synthetic correlated datasets. CoRR Federica Gerace, Luca Saglietti, Stefano Sarao Mannelli, Andrew M. Saxe, Lenka Zdeborová
2020	A*	`conf`	Characterizing emergent representations in a space of candidate learning rules for deep networks. NeurIPS Yinan Cao, Christopher Summerfield, Andrew M. Saxe
2020	J	`jnl`	High-dimensional dynamics of generalization error in neural networks. Neural Networks Madhu S. Advani, Andrew M. Saxe, Haim Sompolinsky
2019	A*	`conf`	Dynamics of stochastic gradient descent for two-layer neural networks in the teacher-student setup. NeurIPS Sebastian Goldt, Madhu Advani, Andrew M. Saxe, Florent Krzakala, Lenka Zdeborová
2019	J	`jnl`	Dynamics of stochastic gradient descent for two-layer neural networks in the teacher-student setup. CoRR Sebastian Goldt, Madhu S. Advani, Andrew M. Saxe, Florent Krzakala, Lenka Zdeborová
2019	J	`jnl`	Generalisation dynamics of online learning in over-parameterised neural networks. CoRR Sebastian Goldt, Madhu S. Advani, Andrew M. Saxe, Florent Krzakala, Lenka Zdeborová
2018	J	`jnl`	A mathematical theory of semantic development in deep neural networks. CoRR Andrew M. Saxe, James L. McClelland, Surya Ganguli
2018	J	`jnl`	Energy-entropy competition and the effectiveness of stochastic gradient descent in machine learning. CoRR Yao Zhang, Andrew M. Saxe, Madhu S. Advani, Alpha A. Lee
2018	—	`conf`	Hierarchical Subtask Discovery with Non-Negative Matrix Factorization. ICLR (Poster) Adam Christopher Earle, Andrew M. Saxe, Benjamin Rosman
2018	J	`jnl`	Minnorm training: an algorithm for training over-parameterized deep neural networks. CoRR Yamini Bansal, Madhu Advani, David D. Cox, Andrew M. Saxe
2018	—	`conf`	On the Information Bottleneck Theory of Deep Learning. ICLR (Poster) Andrew M. Saxe, Yamini Bansal, Joel Dapello, Madhu Advani, Artemy Kolchinsky, Brendan D. Tracey, David D. Cox
2017	J	`jnl`	Hierarchical Subtask Discovery With Non-Negative Matrix Factorization. CoRR Adam Christopher Earle, Andrew M. Saxe, Benjamin Rosman
2017	A*	`conf`	Hierarchy Through Composition with Multitask LMDPs. ICML Andrew M. Saxe, Adam Christopher Earle, Benjamin Rosman
2017	J	`jnl`	High-dimensional dynamics of generalization error in neural networks. CoRR Madhu S. Advani, Andrew M. Saxe
2016	J	`jnl`	Active Long Term Memory Networks. CoRR Tommaso Furlanello, Jiaping Zhao, Andrew M. Saxe, Laurent Itti, Bosco S. Tjan
2016	J	`jnl`	Hierarchy through Composition with Linearly Solvable Markov Decision Processes. CoRR Andrew M. Saxe, Adam Christopher Earle, Benjamin Rosman
2016	—	`conf`	Tensor Switching Networks. NIPS Chuan-Yung Tsai, Andrew M. Saxe, David D. Cox
2016	J	`jnl`	Tensor Switching Networks. CoRR Chuan-Yung Tsai, Andrew M. Saxe, David D. Cox
2016	B	`conf`	Tutorial Workshop on Contemporary Deep Neural Network Models. CogSci James L. McClelland, Steven Stenberg Hansen, Andrew M. Saxe
2014	B	`conf`	Deep Learning and the Brain. CogSci Andrew M. Saxe
2014	—	`conf`	Exact solutions to the nonlinear dynamics of learning in deep linear neural networks. ICLR Andrew M. Saxe, James L. McClelland, Surya Ganguli
2014	B	`conf`	Modeling Perceptual Learning with Deep Networks. CogSci Rachel Lee, Andrew M. Saxe
2014	B	`conf`	Multitask model-free reinforcement learning. CogSci Andrew M. Saxe
2013	B	`conf`	Learning hierarchical categories in deep neural networks. CogSci Andrew M. Saxe, James L. McClelland, Surya Ganguli
2011	A*	`conf`	On Random Weights and Unsupervised Feature Learning. ICML Andrew M. Saxe, Pang Wei Koh, Zhenghao Chen, Maneesh Bhand, Bipin Suresh, Andrew Y. Ng
2011	—	`conf`	Unsupervised learning models of primary cortical receptive fields and receptive field plasticity. NIPS Andrew M. Saxe, Maneesh Bhand, Ritvik Mudur, Bipin Suresh, Andrew Y. Ng
2009	—	`conf`	Measuring Invariances in Deep Networks. NIPS Ian J. Goodfellow, Quoc V. Le, Andrew M. Saxe, Honglak Lee, Andrew Y. Ng
2006	J	`jnl`	Prospect Eleven: Princeton University's entry in the 2005 DARPA Grand Challenge. J. Field Robotics Anand R. Atreya, Bryan C. Cattle, Brendan M. Collins, Benjamin Essenburg, Gordon H. Franken, Andrew M. Saxe, Scott N. Schiffres, Alain L. Kornhauser