Retrievers

All registered Retriever components in DashAI.

Name	Description
AgenticEnsembleRetriever	Consults LLM-chosen children across rounds to build a sufficient pool.
AgenticRerankerRetriever	Selects the best top_k candidates from a child with an LLM.
AgenticSingleRetriever	Iteratively deepens a single child's retrieval under LLM supervision.
BM25Retriever	Sparse retriever using BM25 (Okapi) ranking for document retrieval.
DenseEmbeddingRetriever	Dense retriever that accepts any text embedding converter.
MMRRerankerRetriever	Re-ranks retrieval results using Maximum Marginal Relevance for diversity.
ParallelRetriever	Queries multiple retrievers in parallel and merges their results.
SentenceTransformerCrossEncoderRetriever	Cross-encoder re-ranker powered by SentenceTransformer models.
SequentialRetriever	Queries multiple retrievers in sequence, re-ranking at each step.
TFIDFRetriever	Sparse retriever using TF-IDF vectorization for document retrieval.