Add cross-link from Overview to What is Core AI section

Author: lizzyjunedavis

docs/index.md

@@ -4,7 +4,7 @@ Bring PyTorch models to Core AI for on-device execution.
 
 ## Overview
 
-Core AI PyTorch Extensions (`coreai-torch`) is a Python package that bridges PyTorch and Core AI. You can use it to bring up an existing PyTorch model — exported as a `torch.export.ExportedProgram` — into a Core AI `AIProgram` ready to run on Apple hardware, traversing the FX graph node-by-node and mapping ATen operators to Core AI operations. You can equally use it to author Core AI models directly from PyTorch by composing the library of composite ops in `coreai_torch.composite_ops`, authoring new ops via `register_torch_lowering`, and authoring inline Metal GPU kernels through `TorchMetalKernel` and `register_custom_kernels` — all expressed as PyTorch `nn.Module`s and lowered to Core AI IR that the compiler recognizes and optimizes natively.
+Core AI PyTorch Extensions (`coreai-torch`) is a Python package that bridges PyTorch and Core AI. You can use it to bring up an existing PyTorch model — exported as a `torch.export.ExportedProgram` — into a Core AI `AIProgram` ready to run on Apple hardware, traversing the FX graph node-by-node and mapping ATen operators to Core AI operations. You can equally use it to author Core AI models directly from PyTorch by composing the library of composite ops in `coreai_torch.composite_ops`, authoring new ops via `register_torch_lowering`, and authoring inline Metal GPU kernels through `TorchMetalKernel` and `register_custom_kernels` — all expressed as PyTorch `nn.Module`s and lowered to Core AI IR that the compiler recognizes and optimizes natively. For an overview of the Core AI ecosystem and how coreai-torch fits in, see [What is Core AI?](#what-is-core-ai).
 
 The bring-up pipeline has three steps. First, export your PyTorch model with `torch.export.export` to capture the computation graph. Second, decompose the exported program with `get_decomp_table()`, which lowers composite ATen ops to the primitive set that `TorchConverter` can map while preserving the operations that `TorchConverter` lowers as composite ops. Third, call `TorchConverter().add_exported_program(ep).to_coreai()` to produce the `AIProgram`.