Prerequisites for CLIP, OpenCLIP, and SigLIP: Contrastive Language-Image Pretraining

Direct prerequisites (3)

1. Contrastive Learninglayer 3, tier 2
2. Vision Transformer Lineage: ViT, DeiT, Swin, MAE, DINOv2, SAMlayer 4, tier 1
3. Information Theory Foundationslayer 0B, tier 2

Reachable through the chain (19)

These topics are not directly cited as prerequisites but are reached transitively by following the chain upward. Working through the direct prerequisites pulls these in.

• Feedforward Networks and Backpropagationlayer 2, tier 1
• Differentiation in Rnlayer 0A, tier 1
• Sets, Functions, and Relationslayer 0A, tier 1
• Basic Logic and Proof Techniqueslayer 0A, tier 2
• Vectors, Matrices, and Linear Mapslayer 0A, tier 1
• Continuity in Rⁿlayer 0A, tier 1
• Metric Spaces, Convergence, and Completenesslayer 0A, tier 1
• Matrix Calculuslayer 1, tier 1
• The Jacobian Matrixlayer 0A, tier 1
• The Hessian Matrixlayer 0A, tier 1
• Matrix Operations and Propertieslayer 0A, tier 1
• Eigenvalues and Eigenvectorslayer 0A, tier 1
• Activation Functionslayer 1, tier 1
• Convex Optimization Basicslayer 1, tier 1
• Transformer Architecturelayer 4, tier 2
• Attention Mechanism Theorylayer 4, tier 2
• Softmax and Numerical Stabilitylayer 1, tier 1
• Convolutional Neural Networkslayer 3, tier 2
• Self-Supervised Visionlayer 4, tier 2