2024: Published 'Triple-Encoders: Representations That Fire Together, Wire Together'.
2023: Released DAPR benchmark for Document-Aware Passage Retrieval (arXiv).
2022: Published 'Incorporating Relevance Feedback...' at EMNLP; led the MTEB (Massive Text Embedding Benchmark) project (arXiv); contributed to NeurIPS ENLSP workshop paper 'Efficient Few-Shot Learning Without Prompts'; published 'Domain Adaptation for Memory-Efficient Dense Retrieval' (arXiv); co-authored ACL 2022 demo paper 'UKP-SQUARE'; authored Medium post on GPT-3 text embeddings.
2021: Proposed GPL (Generative Pseudo Labeling) for unsupervised domain adaptation in dense retrieval (arXiv); released BEIR benchmark for zero-shot IR evaluation (NeurIPS 2021); published TWEAC (arXiv), TSDAE (EMNLP Findings), and cooperative cross-modal retrieval approach (TACL 2021).
2020–2021: Published 'The Curse of Dense Low-Dimensional Information Retrieval...' (ACL 2021); published cross-document event coreference resolution in Computational Linguistics journal; co-developed AdapterDrop for efficient adapter usage in Transformers.