Extending Multi-Positive Contrastive Learning for Robust Dense Retrieval Against Query Perturbations

Dense retrieval has become the new paradigm in passage retrieval. Despite its effectiveness on typo-free queries, it is not robust when dealing with queries that contain typos. Current works on improving the typo-robustness of dense retrievers combine (i) data augmentation to obtain the typoed queries during training time with (ii) additional robustifying subtasks that aim to align the original, typo-free queries with their typoed variants. Even though multiple typoed variants are available as positive samples per query, some methods assume a single positive sample and a set of negative ones p

Research goal: Can the multi-positive contrastive learning method be extended to improve robustness against other types of query perturbations (e.g., paraphrasing or synonym substitutions) in dense retrieval, and how does this impact recall@10 on benchmark datasets like TriviaQA or HotpotQA?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 8.5/10.