mlsanityData to 7 Oct 2026

Your Prompt Should Do More: Effects of Retrieval Instructions in Embedding Models

arXiv 2610.10508 ↗First posted 7 Oct 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

Prompted embedding models have recently received increasing attention, particularly for retrieval, where detailed retrieval instructions are provided as part of the retrieval prompt. Several new datasets and studies have examined this setting, showing that the current embedding models often struggle to follow such instructions reliably. In this paper, we study the mechanism of how instructions actually affect the representations of retrieval queries in asymmetric retrieval tasks. We show that models can fail to follow even simple task instructions when query-side distractors are included in the evaluation. We hypothesize that this behavior is driven by the training setup of current embedding models and their evaluation, and show that fine-tuning with added query-side distractors leads to substantial improvements, with minimal effect on other tasks.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 13 of its 41 references that are in the map. First posted on arXiv 7 Oct 2026.

Builds on

13 of 41 references.

References · 13

  1. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,049
  2. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,710
  3. 2025Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Modelscited by 1,175
  4. 2024Multilingual E5 Text Embeddings: A Technical Reportcited by 357
  5. 2024SWIFT:A Scalable lightWeight Infrastructure for Fine-Tuningcited by 345
  6. 2025EmbeddingGemma: Powerful and Lightweight Text Representationscited by 128
  7. 2024NV-Retriever: Improving text embedding models with effective hard-negative miningcited by 83
  8. 2024RAR-b: Reasoning as Retrieval Benchmarkcited by 33
  9. 2024INSTRUCTIR: A Benchmark for Instruction Following of Information Retrieval Modelscited by 32
  10. 2024SyNeg: LLM-Driven Synthetic Hard-Negatives for Dense Retrievalcited by 14
  11. 2026F2LLM-v2: Inclusive, Performant, and Efficient Embeddings for a Multilingual Worldcited by 10
  12. 2025Towards Better Instruction Following Retrieval Modelscited by 9
1 more
  1. 2025Do Prompts Reshape Representations? An Empirical Study of Prompting Effects on Embeddingscited by 1

Built on by

Not cited yet.