Search Results - "Raghu, Maithra"

Insights from Deep Representations for Machine Learning Systems and Human Collaborations by Raghu, Maithra

“…Over the past several years, we have witnessed fundamental breakthroughs in machine learning, largely driven by rapid advancements of the underlying deep…”
Get full text

A Survey of Deep Learning for Scientific Discovery by Raghu, Maithra, Schmidt, Eric

“…Over the past few years, we have seen fundamental breakthroughs in core problems in machine learning, largely driven by advances in deep neural networks. At…”
Get full text

On the Origins of the Block Structure Phenomenon in Neural Network Representations by Nguyen, Thao, Raghu, Maithra, Kornblith, Simon

“…Recent work has uncovered a striking phenomenon in large-capacity neural networks: they contain blocks of contiguous hidden layers with highly similar…”
Get full text

Do Wide and Deep Networks Learn the Same Things? Uncovering How Neural Network Representations Vary with Width and Depth by Nguyen, Thao, Raghu, Maithra, Kornblith, Simon

“…A key factor in the success of deep neural networks is the ability to scale models to improve performance by varying the architecture depth and width. This…”
Get full text

Anatomy of Catastrophic Forgetting: Hidden Representations and Task Semantics by Ramasesh, Vinay V, Dyer, Ethan, Raghu, Maithra

“…A central challenge in developing versatile machine learning systems is catastrophic forgetting: a model trained on tasks in sequence will suffer significant…”
Get full text

Do Vision Transformers See Like Convolutional Neural Networks? by Raghu, Maithra, Unterthiner, Thomas, Kornblith, Simon, Zhang, Chiyuan, Dosovitskiy, Alexey

“…Convolutional neural networks (CNNs) have so far been the de-facto model for visual data. Recent work has shown that (Vision) Transformer models (ViT) can…”
Get full text

Pointer Value Retrieval: A new benchmark for understanding the limits of neural network generalization by Zhang, Chiyuan, Raghu, Maithra, Kleinberg, Jon, Bengio, Samy

“…Central to the success of artificial neural networks is their ability to generalize. But does neural network generalization primarily rely on seeing highly…”
Get full text

Teaching with Commentaries by Raghu, Aniruddh, Raghu, Maithra, Kornblith, Simon, Duvenaud, David, Hinton, Geoffrey

“…Effective training of deep neural networks can be challenging, and there remain many open questions on how to best learn these models. Recently developed…”
Get full text

Learning to Reason with Neural Networks: Generalization, Unseen Data and Boolean Measures by Abbe, Emmanuel, Bengio, Samy, Cornacchia, Elisabetta, Kleinberg, Jon, Lotfi, Aryo, Raghu, Maithra, Zhang, Chiyuan

“…This paper considers the Pointer Value Retrieval (PVR) benchmark introduced in [ZRKB21], where a 'reasoning' function acts on a string of digits to produce the…”
Get full text

Team Performance with Test Scores by Kleinberg, Jon, Raghu, Maithra

“…Team performance is a ubiquitous area of inquiry in the social sciences, and it motivates the problem of team selection -- choosing the members of a team for…”
Get full text

Insights on representational similarity in neural networks with canonical correlation by Morcos, Ari S, Raghu, Maithra, Bengio, Samy

“…Comparing different neural network representations and determining how representations evolve over time remain challenging open questions in our understanding…”
Get full text

Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML by Raghu, Aniruddh, Raghu, Maithra, Bengio, Samy, Vinyals, Oriol

“…An important research direction in machine learning has centered around developing meta-learning algorithms to tackle few-shot learning. An especially…”
Get full text

Transfusion: Understanding Transfer Learning for Medical Imaging by Raghu, Maithra, Zhang, Chiyuan, Kleinberg, Jon, Bengio, Samy

“…Transfer learning from natural image datasets, particularly ImageNet, using standard large models and corresponding pretrained weights has become a de-facto…”
Get full text

The Algorithmic Automation Problem: Prediction, Triage, and Human Effort by Raghu, Maithra, Blumer, Katy, Corrado, Greg, Kleinberg, Jon, Obermeyer, Ziad, Mullainathan, Sendhil

“…In a wide array of areas, algorithms are matching and surpassing the performance of human experts, leading to consideration of the roles of human judgment and…”
Get full text

SVCCA: Singular Vector Canonical Correlation Analysis for Deep Learning Dynamics and Interpretability by Raghu, Maithra, Gilmer, Justin, Yosinski, Jason, Sohl-Dickstein, Jascha

“…We propose a new technique, Singular Vector Canonical Correlation Analysis (SVCCA), a tool for quickly comparing two representations in a way that is both…”
Get full text

Linear Additive Markov Processes by Kumar, Ravi, Raghu, Maithra, Sarlos, Tamas, Tomkins, Andrew

“…We introduce LAMP: the Linear Additive Markov Process. Transitions in LAMP may be influenced by states visited in the distant history of the process, but…”
Get full text

Direct Uncertainty Prediction for Medical Second Opinions by Raghu, Maithra, Blumer, Katy, Sayres, Rory, Obermeyer, Ziad, Kleinberg, Robert, Mullainathan, Sendhil, Kleinberg, Jon

“…The issue of disagreements amongst human experts is a ubiquitous one in both machine learning and medicine. In medicine, this often corresponds to doctor…”
Get full text

Adversarial Spheres by Gilmer, Justin, Metz, Luke, Faghri, Fartash, Schoenholz, Samuel S, Raghu, Maithra, Wattenberg, Martin, Goodfellow, Ian

“…State of the art computer vision models have been shown to be vulnerable to small adversarial perturbations of the input. In other words, most images in the…”
Get full text

Can Deep Reinforcement Learning Solve Erdos-Selfridge-Spencer Games? by Raghu, Maithra, Irpan, Alex, Andreas, Jacob, Kleinberg, Robert, Le, Quoc V, Kleinberg, Jon

“…Deep reinforcement learning has achieved many recent successes, but our understanding of its strengths and limitations is hampered by the lack of rich…”
Get full text

Survey of Expressivity in Deep Neural Networks by Raghu, Maithra, Poole, Ben, Kleinberg, Jon, Ganguli, Surya, Sohl-Dickstein, Jascha

“…We survey results on neural network expressivity described in "On the Expressive Power of Deep Neural Networks". The paper motivates and develops three natural…”
Get full text