Are Convolutional Networks Inherently Foveated?
Bilal Alsallakh, Vivek Miglani, Narine Kokhlikyan, David Adkins, Orion Reblitz-Richardson
People
Orion Reblitz-Richardson
Applied Research Scientist
I am an Applied Research Scientist working on augmented reality and mobile machine learning at Facebook since 2016. Before joining Facebook, I was a Founding Engineer at Will Wright’s Stupid Fun Club doing work on clustering and visualization of media in 3D worlds. Prior to that I founded two social platform startups in Boston and, early in my career, worked on voice recognition at Tellme Networks.
I received my B.S. and M.Eng. in Computer Science from MIT focused on AI. Research included information retrieval and search with professor David Karger, biological modeling with professor C. Forbes Dewey, and symbolic knowledge representation with professor Marvin Minsky in his Society of Mind group.
Interests
Artificial Intelligence, Convolutional Networks, Concept Learning, Unsupervised Learning, General Intelligence, Speech Recognition, and Computer Vision
Latest Publications
Shared Visual Representations in Human and Machine Intelligence (SVRHM) Workshop at NeurIPS - 2021
Debugging the Internals of Convolutional Networks
Bilal Alsallakh, Narine Kokhlikyan, Vivek Miglani, Shubham Muttepawar, Edward Wang, Sara Zhang, David Adkins, Orion Reblitz-Richardson
Captum: a Unified and Generic Model Interpretability Library for PyTorch
Narine Kokhlikyan, Vivek Miglani, Miguel Martin, Edward Wang, Bilal Alsallakh, Jonathan Reynolds, Alexander Melnikov, Natalia Kliushkina, Carlos Araya, Siqi Yan, Orion Reblitz-Richardson
Investigating Effects of Saturation in Integrated Gradients
Vivek Miglani, Bilal Alsallakh, Narine Kokhlikyan, Orion Reblitz-Richardson
