What is the pgvector maximum supported dimensions limit and how do you handle high-dimensional vectors?

-- Check current dimension limit at runtime:
SELECT current_setting('vector.max_dimensions');

-- If you need > 16,000 dimensions (rare in practice):
-- Option 1: Use sparsevec if vector is sparse
CREATE TABLE items (
    id        BIGSERIAL PRIMARY KEY,
    embedding SPARSEVEC(100000)  -- up to 1M dims
);

-- Option 2: Dimensionality reduction before storage
-- Use PCA or UMAP to reduce from e.g. 65536-dim to 2048-dim
-- then store as vector(2048)

-- Option 3: Split vector across columns (workaround, ugly)
-- Not recommended but possible for extreme edge cases

-- Real-world dimension reference:
-- text-embedding-3-small: 1,536 dims (well within limit)
-- text-embedding-3-large: 3,072 dims (well within limit)
-- CLIP image embeddings: 512-768 dims (well within limit)
-- Custom deep models: may go up to 4096 dims
-- All well within the 16,000 dim limit for vector(n)