#ResNet
58 related articles
·4 min
Why Do ResNet Skip Connections Work? Reproducing the Deep Network Degradation Problem
Reproducing the deep network degradation problem on CIFAR-10: a 56-layer plain network achieves only 84% training accuracy vs. 95% for 20 layers. How ResNet skip connections solve this.
Read more →
·3 min
Choosing a Laptop for AI Studies: MacBook vs NVIDIA Laptop — An In-Depth Comparison Guide
In-depth analysis for AI students choosing laptops: MacBook Air M5 with remote GPU vs NVIDIA laptop, comparing CUDA support, portability, battery life, and value.
Read more →
·3 min
Getting Started in Machine Learning Research: Essential Paper Reading List and Research Internship Application Path
A complete path from zero to research internship for ML beginners, covering essential classic papers (AlexNet, ResNet, Transformer), paper reading methods, reproduction tips, and practical advice for research internship applications.
Read more →
·4 min
Learn All the Math for Machine Learning at Zero Cost: A Complete Khan Academy Roadmap
A free ML math learning roadmap based on Khan Academy videos, covering linear algebra, calculus, and probability across nine stages with clear must-learn, optional, and skippable content labels.
Read more →
·3 min
How to Draw CNN Neural Network Architecture Diagrams? A Complete Guide to Tools and Methods
A systematic guide to drawing professional CNN architecture diagrams using mainstream tools like NN-SVG, PlotNeuralNet, Netron, and torchviz for papers and projects.
Read more →
·5 min
From AI to Large Models: Understanding the Conceptual Landscape and Technological Evolution of Artificial Intelligence
Understand how AI, machine learning, deep learning, large models, and generative AI relate to each other. From Deep Blue to ChatGPT, learn how Transformer architecture gave rise to LLMs.
Read more →
·5 min
How Undergraduates Without Research Advisors Can Start Independent Research
How can undergraduates without advisors or labs conduct independent research? This guide covers paper reproduction, open resources, finding remote mentors, and publishing — a complete path for resource-limited students.
Read more →
·3 min
Building a CNN from Scratch in C++: Should You Keep Polishing or Move to Something New?
A first-year student built a CNN from scratch in C++17 reaching 94% accuracy. This article provides a marginal learning returns framework to help ML beginners decide whether to keep polishing or start new projects.
Read more →
·3 min
Poison-Resistant Concept Anchoring: A New Approach to Defending Against AI Data Poisoning
Deep dive into Poison-Resistant Concept Anchoring, defending against data poisoning via signed anchors and bounded updates. Experiments show 62% poison isolation with 0% false rejection rate.
Read more →
·4 min
Deep Dive into Kimi K3 Architecture: KDA, Stable Latent MoE, and Attention Residuals — Three Core Technologies Explained
Deep dive into Kimi K3's three core architecture technologies: KDA memory management, Stable Latent MoE with 896 experts activating only 16, and Attention Residuals — from math to implementation.
Read more →
·3 min
Reproducing the Deep Network Degradation Problem: Why Residual Connections Save Deep Networks
Reproducing the degradation problem on CIFAR-10: a 56-layer plain network achieves only 84% training accuracy vs 95.1% for 20 layers. ResNet adds just 0.3% parameters but boosts accuracy to 99%.
Read more →
·3 min
MRI Brain Tumor Detection Graduation Project: A Complete Technical Roadmap from Scratch
A complete guide for MRI brain tumor detection graduation projects: medical background, BraTS dataset selection, GAN/diffusion model/Transformer technical routes, Research Gap methodology, and Agent collaboration architecture.
Read more →
·4 min
OpenAI Chief Research Officer Mark Chen: Within Three Years, AI Will Independently Conduct End-to-End Scientific Research
OpenAI CRO Mark Chen shares frontier AI research insights: RL boundaries, why Scaling Laws aren't dead, the o1 reasoning model's origin story, and the bold three-year goal of AI conducting end-to-end scientific research independently.
Read more →
·4 min
Advancing Your Self-Taught Machine Learning Journey: Three Paths Forward After Completing Foundational Projects
After completing MNIST implementation and paper reproduction, how should self-taught ML learners advance? This article outlines three paths: computer vision, NLP, and math foundations.
Read more →
·3 min
Training DETR on Small Datasets: A Practical Guide to Dead Tree Detection from Drones
A deep dive into training DETR on just 5,600 drone images for dead tree detection. Covers pre-training, Deformable DETR variants, parameter reduction, and augmentation strategies for small-dataset object detection.
Read more →
·3 min
CNN Core Mechanisms Explained: From Convolution Principles to the Double Descent Phenomenon
Deep dive into CNN core mechanisms including local connectivity, weight sharing, pooling, receptive fields, Dropout regularization, and the still-unexplained Double Descent phenomenon in deep learning.
Read more →
·3 min
Getting Started with Deep Learning: A Complete Guide to Free Learning Resources and Roadmap
A curated guide to free deep learning resources for ML learners, covering Andrew Ng's courses, CS231n, fast.ai, PyTorch tutorials, and a complete learning roadmap from theory to Kaggle practice.
Read more →
·3 min
Master AI Math with 5 Equations: 20 Free Python Hands-On Projects
A free ML workbook distills core machine learning math into 5 equations with 20 runnable Python projects covering gradient descent, backpropagation, loss functions, and more across NumPy, PyTorch, and XGBoost.
Read more →
·3 min
Spectral Pooling Explained: Rethinking CNN Downsampling from a Frequency Domain Perspective
An in-depth analysis of Spectral Pooling: how DFT-based ideal low-pass filtering in the frequency domain solves max pooling's information loss and aliasing problems, with discussion of computational trade-offs.
Read more →
·3 min
How Transformers Think and Store Facts: A Detailed Guide to Mechanistic Interpretability
Deep dive into Transformer internals: how MLP layers store facts as key-value memories, why high-dimensional near-orthogonality enables millions of concepts, and how attention and MLP layers collaborate.
Read more →