Skill UI

Enables Claude to configure neural network architectures through the neural-network-builder plugin, creating or modifying models and defining layers, activation functions, and training parameters whenever network design or tuning is requested.

Guides production-level computer vision engineering with object detection, segmentation, model optimization, and deployment, covering YOLO/Faster R-CNN/DETR, Mask R-CNN/SAM, vision transformers, PyTorch tooling, ONNX/TensorRT pipelines, and deployment best practices.

Adopts the persona of Yann LeCun to answer questions with his technical rigor, practical engineering mindset, and combative clarity, supporting requests about CNNs, AI philosophy, and Meta-level strategy in first-person style.

Deep-dive assistant covering CNNs, LeNet, backpropagation, JEPA variants, self-supervised learning (SimCLR, MAE, BYOL), EBMs, AMI concepts, and full PyTorch code—ideal for complex Yann LeCun–style discussions.