REDSI: Addressing the Reproducibility and Evaluation Consistency of Differentiable Search Indexing for Document Retrieval
TL;DR
ReDSI is introduced, the first open-source DSI implementation supporting all three original document identifier types, together with a parameterizable and well-documented NQ320K construction pipeline, and conducts extensive experiments under model downscaling, covering retrieval effectiveness, parameter efficiency, training methods and decoding strategies.
Abstract
The differentiable search index (DSI) framework (Tay et al., 2022) has become the de facto baseline for generative retrieval. However, DSI is hard to reproduce: no public implementation covers all three original document identifier types (atomic, naive, semantic), reported results vary widely, and the ubiquitous NQ320K dataset is built from Natural Questions through diverse and underspecified preprocessing. We introduce ReDSI, the first open-source DSI implementation supporting all three identifier types, together with a parameterizable and well-documented NQ320K construction pipeline. Experimentally, we achieve results that are competitive with or stronger than previous DSI baselines. Moreover, we conduct extensive experiments under model downscaling, covering retrieval effectiveness, parameter efficiency, training methods and decoding strategies, opening novel directions for future research.