197 citations · 611 across the 23 of their papers we have counts for
25 papers
UNREAL: Unifying Retrieval and Long-Context with a Single Model
Edan Kinderman, Elad Hoffer, Yochai Blau +4
Long-context inference and Retrieval-Augmented Generation (RAG) handle evidence selection at vastly different scales, from a single long prompt to an entire corpus. We ask whether…
Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning
NVIDIA, :, Aaron Blakeman +572
We introduce Nemotron 3 Ultra, a 550 billion total and 55 billion active parameter Mixture-of-Experts Hybrid Mamba-Attention language model. We pre-trained Nemotron 3 Ultra on 20 t…
Workspace Optimization: How to Train Your Agent
Elad Sarafian, Gal Kaplun, Ron Banner +2
Modern agents built on frontier language models often cannot adapt their weights. What, then, remains trainable? We argue it is the agent's \emph{workspace}, the structured externa…
Retrieval from Within: An Intrinsic Capability of Attention-Based Models
Elad Hoffer, Yochai Blau, Edan Kinderman +3
Retrieval-augmented generation (RAG) typically treats retrieval and generation as separate systems. We ask whether an attention-based encoder-decoder can instead retrieve directly…
Normalized Architectures are Natively 4-Bit
Maxim Fishman, Brian Chmiel, Ron Banner +2
Training large language models at 4-bit precision is critical for efficiency. We show that nGPT, an architecture that constrains weights and hidden representations to the unit hype…
FP4 All the Way: Fully Quantized Training of LLMs
Brian Chmiel, Maxim Fishman, Ron Banner +1
We demonstrate, for the first time, fully quantized training (FQT) of large language models (LLMs) using predominantly 4-bit floating-point (FP4) precision for weights, activations…
EXAQ: Exponent Aware Quantization For LLMs Acceleration
Moran Shkolnik, Maxim Fishman, Brian Chmiel +3
Quantization has established itself as the primary approach for decreasing the computational and storage expenses associated with Large Language Models (LLMs) inference. The majori…
Scaling FP8 training to trillion-token LLMs
Maxim Fishman, Brian Chmiel, Ron Banner +1
We train, for the first time, large language models using FP8 precision on datasets up to 2 trillion tokens -- a 20-fold increase over previous limits. Through these extended train…
DropCompute: simple and more robust distributed synchronous training via compute variance reduction
Niv Giladi, Shahar Gottlieb, Moran Shkolnik +5
Background: Distributed training is essential for large scale training of deep neural networks (DNNs). The dominant methods for large scale DNN training are synchronous (e.g. All-R…
Minimum Variance Unbiased N:M Sparsity for the Neural Gradients
Brian Chmiel, Itay Hubara, Ron Banner +1
In deep learning, fine-grained N:M sparsity reduces the data footprint and bandwidth of a General Matrix multiply (GEMM) up to x2, and doubles throughput by skipping computation of…
Energy awareness in low precision neural networks
Nurit Spingarn Eliezer, Ron Banner, Elad Hoffer +2
Power consumption is a major obstacle in the deployment of deep neural networks (DNNs) on end devices. Existing approaches for reducing power consumption rely on quite general prin…
Graph Representation Learning via Aggregation Enhancement
Maxim Fishman, Chaim Baskin, Evgenii Zheltonozhskii +3
Graph neural networks (GNNs) have become a powerful tool for processing graph-structured data but still face challenges in effectively aggregating and propagating information betwe…
Accurate Neural Training with 4-bit Matrix Multiplications at Standard Formats
Brian Chmiel, Ron Banner, Elad Hoffer +2
Quantization of the weights and activations is one of the main methods to reduce the computational footprint of Deep Neural Networks (DNNs) training. Current methods enable 4-bit q…
Accelerated Sparse Neural Training: A Provable and Efficient Method to Find N:M Transposable Masks
Itay Hubara, Brian Chmiel, Moshe Island +3
Unstructured pruning reduces the memory footprint in deep neural networks (DNNs). Recently, researchers proposed different types of structural pruning intending to reduce also the…
GAN "Steerability" without optimization
Nurit Spingarn-Eliezer, Ron Banner, Tomer Michaeli
Recent research has shown remarkable success in revealing "steering" directions in the latent spaces of pre-trained GANs. These directions correspond to semantically meaningful ima…
Improving Post Training Neural Quantization: Layer-wise Calibration and Integer Programming
Itay Hubara, Yury Nahshan, Yair Hanani +2
Lately, post-training quantization methods have gained considerable attention, as they are simple to use, and require only a small unlabeled calibration set. This small dataset can…
Neural gradients are near-lognormal: improved quantized and sparse training
Brian Chmiel, Liad Ben-Uri, Moran Shkolnik +3
While training can mostly be accelerated by reducing the time needed to propagate neural gradients back throughout the model, most previous works focus on the quantization/pruning…
Robust Quantization: One Model to Rule Them All
Moran Shkolnik, Brian Chmiel, Ron Banner +4
Neural network quantization methods often involve simulating the quantization process during training, making the trained model highly dependent on the target bit-width and precise…
Loss Aware Post-training Quantization
Yury Nahshan, Brian Chmiel, Chaim Baskin +4
Neural network quantization enables the deployment of large models on resource-constrained devices. Current post-training quantization methods fall short in terms of accuracy for I…
CAT: Compression-Aware Training for bandwidth reduction
Chaim Baskin, Brian Chmiel, Evgenii Zheltonozhskii +3
Convolutional neural networks (CNNs) have become the dominant neural network architecture for solving visual processing tasks. One of the major obstacles hindering the ubiquitous u…
Thanks for Nothing: Predicting Zero-Valued Activations with Lightweight Convolutional Neural Networks
Gil Shomron, Ron Banner, Moran Shkolnik +1
Convolutional neural networks (CNNs) introduce state-of-the-art results for various tasks with the price of high computational demands. Inspired by the observation that spatial cor…
Feature Map Transform Coding for Energy-Efficient CNN Inference
Brian Chmiel, Chaim Baskin, Ron Banner +5
Convolutional neural networks (CNNs) achieve state-of-the-art accuracy in a variety of tasks in computer vision and beyond. One of the major obstacles hindering the ubiquitous use…
Post-training 4-bit quantization of convolution networks for rapid-deployment
Ron Banner, Yury Nahshan, Elad Hoffer +1
Convolutional neural networks require significant memory bandwidth and storage for intermediate computations, apart from substantial computing resources. Neural network quantizatio…
Scalable Methods for 8-bit Training of Neural Networks
Ron Banner, Itay Hubara, Elad Hoffer +1
Quantized Neural Networks (QNNs) are often used to improve network efficiency during the inference phase, i.e. after the network has been trained. Extensive research in the field s…