Publications (28)
Towards Foundation Models for Knowledge Graph Reasoning
Mikhail Galkin, Xinyu Yuan, Hesham Mostafa +2
Foundation models in language and vision have the ability to run inference on any textual and visual inputs thanks to the transferable representations such as a vocabulary of token…
Synaptic Plasticity Dynamics for Deep Continuous Local Learning (DECOLLE)
Jacques Kaiser, Hesham Mostafa, Emre Neftci
A growing body of work underlines striking similarities between biological neural networks and recurrent, binary neural networks. A relatively smaller body of work, however, discus…
NullHop: A Flexible Convolutional Neural Network Accelerator Based on Sparse Representations of Feature Maps
Alessandro Aimar, Hesham Mostafa, Enrico Calabrese +8
Convolutional neural networks (CNNs) have become the dominant neural network architecture for solving many state-of-the-art (SOA) visual processing tasks. Even though Graphical Pro…
PARSAC: Fast, Human-quality Floorplanning for Modern SoCs with Complex Design Constraints
Hesham Mostafa, Uday Mallappa, Mikhail Galkin +2
The floorplanning of Systems-on-a-Chip (SoCs) and of chip sub-systems is a crucial step in the physical design flow as it determines the optimal shapes and locations of the blocks…
Distributed Training of Large Graph Neural Networks with Variable Communication Rates
Juan Cervino, Md Asadullah Turja, Hesham Mostafa +2
Training Graph Neural Networks (GNNs) on large graphs presents unique challenges due to the large memory and computing requirements. Distributed GNN training, where the graph is pa…
Sequential Aggregation and Rematerialization: Distributed Full-batch Training of Graph Neural Networks on Large Graphs
Hesham Mostafa
We present the Sequential Aggregation and Rematerialization (SAR) scheme for distributed full-batch training of Graph Neural Networks (GNNs) on large graphs. Large-scale training o…
Permutohedral-GCN: Graph Convolutional Networks with Global Attention
Hesham Mostafa, Marcel Nassar
Graph convolutional networks (GCNs) update a node's feature vector by aggregating features from its neighbors in the graph. This ignores potentially useful contributions from dista…
DiRotQ: Rotation-Aware Quantization for 4-bit Diffusion Transformers
Sayeh Sharify, Mahsa Salmani, Hesham Mostafa
Diffusion Transformers (DiTs) achieve state-of-the-art image generation quality but incur substantial memory and computational costs at inference. While aggressive Post-Training Qu…
A learning framework for winner-take-all networks with stochastic synapses
Hesham Mostafa, Gert Cauwenberghs
Many recent generative models make use of neural networks to transform the probability distribution of a simple low-dimensional noise process into the complex distribution of the d…
On Local Aggregation in Heterophilic Graphs
Hesham Mostafa, Marcel Nassar, Somdeb Majumdar
Many recent works have studied the performance of Graph Neural Networks (GNNs) in the context of graph homophily - a label-dependent measure of connectivity. Traditional GNNs gener…
Early Attentive Sparsification Accelerates Neural Speech Transcription
Zifei Xu, Sayeh Sharify, Hesham Mostafa +3
Transformer-based neural speech processing has achieved state-of-the-art performance. Since speech audio signals are known to be highly compressible, here we seek to accelerate neu…
Fully-inductive Node Classification on Arbitrary Graphs
Jianan Zhao, Zhaocheng Zhu, Mikhail Galkin +3
One fundamental challenge in graph machine learning is generalizing to new graphs. Many existing methods following the inductive setup can generalize to test graphs with new struct…
Exploiting Long-Term Dependencies for Generating Dynamic Scene Graphs
Shengyu Feng, Subarna Tripathi, Hesham Mostafa +2
Dynamic scene graph generation from a video is challenging due to the temporal dynamics of the scene and the inherent temporal fluctuations of predictions. We hypothesize that capt…
Single-bit-per-weight deep convolutional neural networks without batch-normalization layers for embedded systems
Mark D. McDonnell, Hesham Mostafa, Runchun Wang +1
Batch-normalization (BN) layers are thought to be an integrally important layer type in today's state-of-the-art deep convolutional neural networks for computer vision tasks such a…
Supervised learning based on temporal coding in spiking neural networks
Hesham Mostafa
Gradient descent training techniques are remarkably successful in training analog-valued artificial neural networks (ANNs). Such training techniques, however, do not transfer easil…
Attention-based Image Upsampling
Souvik Kundu, Hesham Mostafa, Sharath Nittur Sridhar +1
Convolutional layers are an integral part of many deep neural network solutions in computer vision. Recent work shows that replacing the standard convolution operation with mechani…
FastSample: Accelerating Distributed Graph Neural Network Training for Billion-Scale Graphs
Hesham Mostafa, Adam Grabowski, Md Asadullah Turja +3
Training Graph Neural Networks(GNNs) on a large monolithic graph presents unique challenges as the graph cannot fit within a single machine and it cannot be decomposed into smaller…
Parameter Efficient Training of Deep Convolutional Neural Networks by Dynamic Sparse Reparameterization
Hesham Mostafa, Xin Wang
Modern deep neural networks are typically highly overparameterized. Pruning techniques are able to remove a significant fraction of network parameters with little loss in accuracy.…
Rhythmic inhibition allows neural networks to search for maximally consistent states
Hesham Mostafa, Lorenz K. Muller, Giacomo Indiveri
Gamma-band rhythmic inhibition is a ubiquitous phenomenon in neural circuits yet its computational role still remains elusive. We show that a model of Gamma-band rhythmic inhibitio…
Stochastic Interpretation of Quasi-periodic Event-based Systems
Hesham Mostafa, Giacomo Indiveri
Many networks used in machine learning and as models of biological neural networks make use of stochastic neurons or neuron-like units. We show that stochastic artificial neurons c…
FloorSet -- a VLSI Floorplanning Dataset with Design Constraints of Real-World SoCs
Uday Mallappa, Hesham Mostafa, Mikhail Galkin +2
Floorplanning for systems-on-a-chip (SoCs) and its sub-systems is a crucial and non-trivial step of the physical design flow. It represents a difficult combinatorial optimization p…
An event-based architecture for solving constraint satisfaction problems
Hesham Mostafa, Lorenz K. Müller, Giacomo Indiveri
Constraint satisfaction problems (CSPs) are typically solved using conventional von Neumann computing architectures. However, these architectures do not reflect the distributed nat…
MF-QAT: Multi-Format Quantization-Aware Training for Elastic Inference
Zifei Xu, Sayeh Sharify, Hesham Mostafa
Quantization-aware training (QAT) is typically performed for a single target numeric format, while practical deployments often need to choose numerical precision at inference time…
Implicit SVD for Graph Representation Learning
Sami Abu-El-Haija, Hesham Mostafa, Marcel Nassar +3
Recent improvements in the performance of state-of-the-art (SOTA) methods for Graph Representational Learning (GRL) have come at the cost of significant computational resource requ…
Hardware-efficient on-line learning through pipelined truncated-error backpropagation in binary-state networks
Hesham Mostafa, Bruno Pedroni, Sadique Sheik +1
Artificial neural networks (ANNs) trained using backpropagation are powerful learning architectures that have achieved state-of-the-art performance in various benchmarks. Significa…
Surrogate Gradient Learning in Spiking Neural Networks
Emre O. Neftci, Hesham Mostafa, Friedemann Zenke
Spiking neural networks are nature's versatile solution to fault-tolerant and energy efficient signal processing. To translate these benefits into hardware, a growing number of neu…
Robust Federated Learning Through Representation Matching and Adaptive Hyper-parameters
Hesham Mostafa
Federated learning is a distributed, privacy-aware learning scenario which trains a single model on data belonging to several clients. Each client trains a local model on its data…
Deep supervised learning using local errors
Hesham Mostafa, Vishwajith Ramesh, Gert Cauwenberghs
Error backpropagation is a highly effective mechanism for learning high-quality hierarchical features in deep networks. Updating the features or weights in one layer, however, requ…