Sequence-to-Sequence

• Sequence-to-Sequence

  Encoder-decoder architecture mapping input sequences to output sequences; used for translation, summarisation, and dialogue.

• DocT5Query

  Document expansion via T5 query generation; generates synthetic queries a document might answer and appends them to the document before indexing, improving sparse retrieval recall.

• DSI (Differentiable Search Index)

  Encodes an entire document corpus into a single seq2seq model; retrieval is performed by generating document identifiers directly from a query, without a separate index.

• FiD (Fusion-in-Decoder)

  Encodes each retrieved passage independently with T5, then fuses all passage representations in the decoder; more scalable than concatenating all passages as a single long input.

• MonoT5

  T5-based pointwise reranker that generates “true”/“false” tokens to score relevance; more efficient than MonoBERT and generalizes well across domains.

• RankT5

  T5-based listwise reranker that directly optimizes ranking metrics by generating ordered document IDs; addresses exposure bias in pointwise and pairwise approaches.