Free Information Center
AI for drug discovery refers to the application of artificial intelligence technologies to enhance and accelerate the process of identifying, designing, and developing new pharmaceutical compounds. By leveraging machine learning, deep learning, and other AI methods, researchers aim to improve the efficiency, accuracy, and cost-effectiveness of drug development.
VoxPoser integrates language conditions into robotic manipulation, enhancing human-robot interactions.
In-processing bias mitigation refers to methods employed during data processing to reduce bias in machine learning and AI systems.
A liquid state machine is a theoretical model in artificial intelligence that processes information in a fluid-like manner, allowing for dynamic adaptation.
Ernie is a series of language models developed by Baidu that utilize knowledge-enhanced pre-training techniques to improve natural language understanding and generation tasks. It incorporates structured knowledge into the training process to enhance performance on various AI benchmarks.
OpenFold is an open-source implementation of DeepMind’s AlphaFold protein structure prediction model. It aims to provide accessible tools for protein folding research and development by replicating and extending AlphaFold’s capabilities using public resources.
Gradient masking is a technique used in machine learning to obscure or alter gradient information, often employed in adversarial machine learning to prevent model vulnerabilities from being exploited. It affects the ability of attackers to generate effective adversarial examples by hiding or manipulating gradient signals.
Program-aided language model (PoT) is a class of language models that enhance natural language understanding and generation by integrating programmatic reasoning. PoT models leverage symbolic computation or external program modules to improve the interpretability and accuracy of language tasks.
PointCNN is a convolutional neural network architecture designed specifically for processing point cloud data by learning an X-transformation to reorder and weight points before applying convolution.
Point cloud transformers are a class of deep learning models designed to process and analyze 3D point cloud data using transformer architectures. They enable efficient feature extraction and global context modeling for tasks such as 3D object recognition and segmentation.