Tag

#deep learning

11 articles

Sigmoid vs ReLU Activation Functions: The Inference Cost of Losing Geometric Context

This article explains how sigmoid and ReLU activation functions affect geometric context preservation in neural networks, and why this matters for inference accuracy.

Apr 817

Step by Step Guide to Build an End-to-End Model Optimization Pipeline with NVIDIA Model Optimizer Using FastNAS Pruning and Fine-Tuning

Learn to build a complete model optimization pipeline using NVIDIA Model Optimizer with FastNAS pruning and fine-tuning techniques. This beginner-friendly tutorial walks you through training, pruning, and fine-tuning a ResNet model on CIFAR-10 dataset.

Apr 212

Nomadic raises $8.4 million to wrangle the data pouring off autonomous vehicles

Autonomous vehicle data management startup Nomadic has raised $8.4 million to develop AI tools that transform raw video footage into structured, searchable datasets. The company's deep learning models address a critical bottleneck in autonomous vehicle development by converting massive unstructured data streams into actionable insights for developers and manufacturers.

Mar 3127

Implementing Deep Q-Learning (DQN) from Scratch Using RLax JAX Haiku and Optax to Train a CartPole Reinforcement Learning Agent

Learn to build a Deep Q-Network (DQN) reinforcement learning agent from scratch using JAX, RLax, Haiku, and Optax to solve the CartPole environment.

Mar 2239

Math needs thinking time, everyday knowledge needs memory, and a new Transformer architecture aims to deliver both

This article explains how a new AI model uses memory and flexible thinking time to solve problems more efficiently than traditional models.

Mar 2138

Multiverse Computing pushes its compressed AI models into the mainstream

Learn about model compression techniques that reduce the size and computational requirements of large AI models while maintaining performance, enabling broader AI deployment.

Mar 1837

Moonshot AI Releases 𝑨𝒕𝒕𝒆𝒏𝒕𝒊𝒐𝒏 𝑹𝒆𝒔𝒊𝒅𝒖𝒂𝒍𝒔 to Replace Fixed Residual Mixing with Depth-Wise Attention for Better Scaling in Transformers

This article explains how a new AI technique called Attention Residuals changes the way information flows in Transformer models, potentially making them more efficient and easier to train.

Mar 1547

ByteDance reportedly pauses global launch of its Seedance 2.0 video generator

Learn to build a video generation pipeline using Python, Stable Diffusion, and OpenCV, gaining hands-on experience with AI video generation technology similar to what ByteDance's Seedance 2.0 uses.

Mar 1551

RL agents go from face-planting to parkour when researchers keep adding network layers

Learn how adding more layers to AI neural networks helped reinforcement learning agents master complex behaviors like parkour, showing the power of deeper AI systems.

Mar 1533

Adobe Research Unlocking Long-Term Memory in Video World Models with State-Space Models

Learn how State-Space Models (SSMs) help AI systems remember long sequences in videos, solving a major challenge in video generation and understanding.

Feb 2668

Liquid AI’s New LFM2-24B-A2B Hybrid Architecture Blends Attention with Convolutions to Solve the Scaling Bottlenecks of Modern LLMs

Learn to build a hybrid neural network architecture that combines attention mechanisms with convolutional layers, similar to Liquid AI's LFM2-24B-A2B model, to address scaling bottlenecks in large language models.

Feb 2557