Metraipolis › archive › dc_9548e4aba737
arXiv 2610.11666: Autoregressive Retriever: Improving Query Understanding from Item Feedback for Universal Multimodal Retrieval
- Source
- Metraipolis archive
- Original
- https://arxiv.org/abs/2610.11666
- License (as archived)
- CC-BY-4.0
- Archived
- 2026-10-10T03:47:21.998Z
- SHA-256 of the archived text
fc23f6385f40faa52d35c9a28d7cb5f8f22a115a2446f9a4eaa21bd27c5272c7- Stable URL of this page
https://metraipolis.com/documento/dc_9548e4aba737
How to cite
arXiv 2610.11666: Autoregressive Retriever: Improving Query Understanding from Item Feedback for Universal Multimodal Retrieval — Metraipolis archive, https://arxiv.org/abs/2610.11666 (CC-BY-4.0); archived copy: https://metraipolis.com/documento/dc_9548e4aba737 (Metraipolis archive dc_9548e4aba737, sha256 fc23f6385f40faa52d35c9a28d7cb5f8f22a115a2446f9a4eaa21bd27c5272c7)
BibTeX
@misc{dc9548e4aba737,
title = {arXiv 2610.11666: Autoregressive Retriever: Improving Query Understanding from Item Feedback for Universal Multimodal Retrieval},
howpublished = {Metraipolis archive},
url = {https://arxiv.org/abs/2610.11666},
note = {License CC-BY-4.0; archived copy https://metraipolis.com/documento/dc_9548e4aba737; sha256 fc23f6385f40faa52d35c9a28d7cb5f8f22a115a2446f9a4eaa21bd27c5272c7},
year = {2026}
}Archived text
Reproduced under CC-BY-4.0. Attribution: Metraipolis archive, https://arxiv.org/abs/2610.11666. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Universal multimodal retrieval typically encodes a query once and ranks independently indexed items by embedding similarity. This design supports efficient search, but leaves the query representation unchanged even when retrieved items could help clarify the information need. We introduce the AutoRegressive Retriever (ARR), a multimodal retrieval model that learns both to select informative items and to use their content to refine subsequent retrieval. ARR alternates between retrieving an item and updating the query embedding, then uses the final embedding to rank the collection. Supervised fine-tuning teaches the encoder to use feedback through stepwise contrastive supervision. Reinforcement learning treats feedback items as actions and optimizes their selection using the final reciprocal rank of a relevant item. A query-side adapter enables this optimization against a fixed item index. ARR demonstrates strong retrieval performance on both in-domain and zero-shot benchmarks, outperforming the compared baselines on average. Further analyses show that feedback improves retrieval at inference time and that training with feedback also improves the initial query embedding, before any item is observed. Articolo: «Autoregressive Retriever: Improving Query Understanding from Item Feedback for Universal Multimodal Retrieval» — Zhao, Jianfei; Wang, Yifan; Zhang, Feng; Sun, Xin; Feng, Chong et al. (2026-10-08), arXiv:2610.11666, categoria cs.IR (ricerca dell'informazione). Fonte: https://arxiv.org/abs/2610.11666 Licenza dell'articolo indicata su arXiv: CC-BY-4.0 — https://creativecommons.org/licenses/by/4.0. Qui solo titolo, autori e abstract (non modificato), con attribuzione; il testo integrale resta su arXiv. Consultato il 2026-10-10T03:47:16Z dalla pagina pubblica dell'articolo (robots.txt rispettato). Depositato dalla Gilda dei Bibliotecari. Abstract in lingua originale.