Adityanarayanan Radhakrishnan

28 papers A* 3A 1Journal 23

Year	Rank	Type	Title / Venue / Authors
2026	J	`jnl`	Efficient and accurate steering of Large Language Models through attention-guided feature learning. CoRR Parmida Davarmanesh, Ashia Wilson, Adityanarayanan Radhakrishnan
2025	J	`jnl`	Aggregate and conquer: detecting and steering LLM concepts by combining nonlinear predictors over multiple layers. CoRR Daniel Beaglehole, Adityanarayanan Radhakrishnan, Enric Boix-Adserà, Mikhail Belkin
2025	A*	`conf`	Emergence in non-neural models: grokking modular arithmetic via average gradient outer product. ICML Neil Mallinar, Daniel Beaglehole, Libin Zhu, Adityanarayanan Radhakrishnan, Parthe Pandit, Mikhail Belkin
2025	J	`jnl`	xRFM: Accurate, scalable, and interpretable feature learning models for tabular data. CoRR Daniel Beaglehole, David Holzmüller, Adityanarayanan Radhakrishnan, Mikhail Belkin
2024	A*	`conf`	Catapults in SGD: spikes in the training loss and their impact on generalization through feature learning. ICML Libin Zhu, Chaoyue Liu, Adityanarayanan Radhakrishnan, Mikhail Belkin
2024	J	`jnl`	Context-Scaling versus Task-Scaling in In-Context Learning. CoRR Amirhesam Abedsoltan, Adityanarayanan Radhakrishnan, Jingfeng Wu, Mikhail Belkin
2024	J	`jnl`	Emergence in non-neural models: grokking modular arithmetic via average gradient outer product. CoRR Neil Mallinar, Daniel Beaglehole, Libin Zhu, Adityanarayanan Radhakrishnan, Parthe Pandit, Mikhail Belkin
2024	J	`jnl`	Linear Recursive Feature Machines provably recover low-rank matrices. CoRR Adityanarayanan Radhakrishnan, Mikhail Belkin, Dmitriy Drusvyatskiy
2024	A*	`conf`	Quadratic models for understanding catapult dynamics of neural networks. ICLR Libin Zhu, Chaoyue Liu, Adityanarayanan Radhakrishnan, Mikhail Belkin
2023	J	`jnl`	Catapults in SGD: spikes in the training loss and their impact on generalization through feature learning. CoRR Libin Zhu, Chaoyue Liu, Adityanarayanan Radhakrishnan, Mikhail Belkin
2023	—	`—`	Foundations of Machine Learning: Over-parameterization and Feature Learning Adityanarayanan Radhakrishnan
2023	J	`jnl`	Mechanism of feature learning in convolutional neural networks. CoRR Daniel Beaglehole, Adityanarayanan Radhakrishnan, Parthe Pandit, Mikhail Belkin
2022	J	`jnl`	Feature learning in neural networks and kernel machines that recursively learn features. CoRR Adityanarayanan Radhakrishnan, Daniel Beaglehole, Parthe Pandit, Mikhail Belkin
2022	J	`jnl`	Quadratic models for understanding neural network dynamics. CoRR Libin Zhu, Chaoyue Liu, Adityanarayanan Radhakrishnan, Mikhail Belkin
2022	J	`jnl`	Transfer Learning with Kernel Methods. CoRR Adityanarayanan Radhakrishnan, Max Ruiz Luyten, Neha Prasad, Caroline Uhler
2022	J	`jnl`	Wide and Deep Neural Networks Achieve Optimality for Classification. CoRR Adityanarayanan Radhakrishnan, Mikhail Belkin, Caroline Uhler
2021	J	`jnl`	A Mechanism for Producing Aligned Latent Spaces with Autoencoders. CoRR Saachi Jain, Adityanarayanan Radhakrishnan, Caroline Uhler
2021	J	`jnl`	Local Quadratic Convergence of Stochastic Gradient Descent with Adaptive Step Size. CoRR Adityanarayanan Radhakrishnan, Mikhail Belkin, Caroline Uhler
2021	J	`jnl`	Simple, Fast, and Flexible Framework for Matrix Completion with Infinite Width Neural Networks. CoRR Adityanarayanan Radhakrishnan, George Stefanakis, Mikhail Belkin, Caroline Uhler
2020	J	`jnl`	Balancedness and Alignment are Unlikely in Linear Neural Networks. CoRR Adityanarayanan Radhakrishnan, Eshaan Nichani, Daniel Irving Bernstein, Caroline Uhler
2020	J	`jnl`	Do Deeper Convolutional Networks Perform Better? CoRR Eshaan Nichani, Adityanarayanan Radhakrishnan, Caroline Uhler
2020	J	`jnl`	Linear Convergence and Implicit Regularization of Generalized Mirror Descent with Time-Dependent Mirrors. CoRR Adityanarayanan Radhakrishnan, Mikhail Belkin, Caroline Uhler
2020	J	`jnl`	Overparameterized neural networks implement associative memory. Proc. Natl. Acad. Sci. USA Adityanarayanan Radhakrishnan, Mikhail Belkin, Caroline Uhler
2019	J	`jnl`	Overparameterized Neural Networks Can Implement Associative Memory. CoRR Adityanarayanan Radhakrishnan, Mikhail Belkin, Caroline Uhler
2018	J	`jnl`	Counting Markov equivalence classes for DAG models on trees. Discret. Appl. Math. Adityanarayanan Radhakrishnan, Liam Solus, Caroline Uhler
2018	J	`jnl`	Downsampling leads to Image Memorization in Convolutional Autoencoders. CoRR Adityanarayanan Radhakrishnan, Mikhail Belkin, Caroline Uhler
2017	A	`conf`	Counting Markov Equivalence Classes by Number of Immoralities. UAI Adityanarayanan Radhakrishnan, Liam Solus, Caroline Uhler
2017	J	`jnl`	Patchnet: Interpretable Neural Networks for Image Classification. CoRR Adityanarayanan Radhakrishnan, Charles Durham, Ali Soylemezoglu, Caroline Uhler