[Question]: retrieval_by_children

Self Checks

• I have searched for existing issues search for existing issues, including closed ones.
• I confirm that I am using English to submit this report (Language Policy).
• Non-english title submitions will be closed directly ( 非英文标题的提交将会被直接关闭 ) (Language Policy).
• Please do not modify this template :) and fill in all the required fields.

Describe your problem

In the current implementation of retrieval_by_children, all three scoring fields are derived from the same source "similarity": np.mean([ck["similarity"] for ck in cks]), "vector_similarity": np.mean([ck["similarity"] for ck in cks]), "term_similarity": np.mean([ck["similarity"] for ck in cks]), Could you clarify whether this is the intended design?

Specifically:

Is similarity now considered the only canonical scoring signal in the system at this stage? Or was the original intention to preserve separate signals: vector_similarity (embedding-based score) term_similarity (lexical/BM25 score)

and they are currently being overwritten during the aggregation step?