Applied AI·Embeddings and retrieval
you raised k to twenty, the right chunk was definitely in there, and the model missed it among the noise anyway.
Top-k retrieval
Draft summary, pending review
Returning the k best chunks, k typically 3 to 20. Small k risks missing the answer; large k pads the context with noise, cost and lost-in-the-middle risk. Tune it with an eval set, not by feel.