Adapter
A tiny set of highly trainable parameters injected into the layers of a frozen, pre-trained neural network to drastically simplify task-specific tuning.
Think of It Like This
Like adding a custom optical filter to the front of an expensive camera lens instead of rebuilding the entire lens from scratch.
By only updating these inserted bottleneck layers instead of the massive original weights, adapters drastically reduce the memory and compute overhead of fine-tuning. This allows engineers to train dozens of task-specific models that share a single underlying foundation model, swapping the lightweight adapters in and out as needed during inference.