ALLPCB
OpenAI's Defense Against LLM Attacks
OpenAI's study unveils an instruction hierarchy to boost LLM security against attacks like prompt injections, enhancing model safety.
Dynamic Memory Compression for KV Cache During Inference
Dynamic Memory Compression (DMC) compresses the Transformer KV cache online during autoregressive inference, improving throughput and enabling longer context windows.
Latest Trends in Machine Learning Hardware Performance
Analysis of ML hardware trends across GPUs and accelerators, quantifying compute performance, interconnects, cost-performance, and energy efficiency.
GEAR: Efficient KV Cache Compression for LLM Inference
GEAR: hybrid KV cache compression combining 4-bit quantization, low-rank residual approximation, and sparse corrections to cut peak memory and boost inference throughput
Training Large Models Across 10,000+ GPUs
MegaScale system for large-scale LLM training beyond 10,000 GPUs, detailing algorithm-system co-design, communication and network tuning, MFU improvements, and fault-tolerant recovery.
AI Applications in Wargaming: Case Studies
Analysis of AI applications in wargaming: case studies, benefits like enhanced scenario realism and decision support, and limitations such as black-box effects and cost.
Key Steps for Edge AI: Hardware Selection and Model Deployment
Edge AI technical overview covering distributed architecture, model lightweighting, data preprocessing, and model deployment to edge devices for real-time inference.
FPGA Applications in Machine Learning
Overview of FPGA applications in machine learning: accelerating neural network inference, hardware quantization, algorithm optimization, and efficiency for edge AI deployments.
What Is Automatic Speech Recognition and GPU-Accelerated ASR
Overview of ASR (speech-to-text): pipeline, acoustic and language models, CTC training, decoding strategies, and GPU acceleration using NVIDIA NeMo and toolkits.
MAX78000 Low-Power CNN AI Microcontroller
MAX78000 AI microcontroller with low-power CNN accelerator, Arm Cortex?M4 with FPU, 442 KB weight SRAM and 512 KB flash, optimized for edge inference.
Large-Scale Visual SLAM with 3D Gaussian Splatting
LSG-SLAM: a stereo visual SLAM using 3D Gaussian splatting for large-scale outdoor reconstruction, improving tracking stability and mapping quality.
Survey of Deep Metric Learning
Survey of deep metric learning: formulations, sample selection and metric loss functions (contrastive, triplet, N?pair), architectures and applications in vision, audio, and text.
Change Detection in Remote Sensing Using Foundation Models
Time-Traveling Pixels integrates SAM into remote sensing change detection, using low-rank fine-tuning and a Time-Travel Activation Gate to mitigate spatial-semantic domain shift.
Relationship Between AI, Machine Learning, and Deep Learning
Overview of artificial intelligence and its relationship to machine learning and deep learning, covering AI categories, ML workflow, and common deep architectures.
Understanding GRU in Neural Networks
Learn about Gated Recurrent Units (GRU) in neural networks, their role in sequential data processing, and applications in machine learning.
Embodied AI Simulation Generation System
LLM-driven system automates embodied intelligence skill generation for robotic arms, converting natural-language tasks into MuJoCo scenes, actions, and reward code.
Deep Learning for Radar Target Detection
Survey of deep learning approaches for radar target detection, comparing two-stage and single-stage detectors (Faster R-CNN, YOLOv5), preprocessing, and deployment results.
Common Deep Learning Applications in AI
Survey of deep learning applications in AI: image recognition, NLP, speech, recommendation systems, autonomous driving, healthcare, cybersecurity, and VR.
How CNNs Solve Edge Detection
Overview of convolutional neural networks: principles like padding, stride, pooling and filters, edge detection fundamentals, architecture patterns and a Keras MNIST implementation.
Smart Classroom IoT Solution
Analysis of an IoT smart classroom solution - hardware connectivity, data interoperability and scenario intelligence for unified, energy-efficient device management.