Search papers, labs, and topics across Lattice.
This paper introduces a dual-view training strategy for instruction-following information retrieval (IF-IR) that uses polarity reversal to generate complementary instructions for the same document pair, effectively swapping their relevance labels. By forcing the retriever to reconsider candidates based on the instruction rather than just topical cues, the method significantly improves instruction sensitivity. Experiments on the FollowIR benchmark show a 45% performance improvement using a 305M-parameter encoder, outperforming larger general-purpose models.
Flipping relevance labels via LLM-generated complementary instructions boosts instruction-following retrieval by 45%, proving that targeted data synthesis beats brute-force scaling.
Instruction-following information retrieval (IF-IR) studies retrieval systems that must not only find documents relevant to a query, but also obey explicit user constraints such as required attributes, exclusions, or output preferences. However, most retrievers are trained primarily for semantic relevance and often fail to distinguish documents that match the topic from those that satisfy the instruction. We propose a dual-view data synthesis strategy based on polarity reversal: given a query, a document that is relevant under the instruction, and a hard negative that matches the query but violates the instruction, we prompt an LLM to generate a complementary instruction under which the two documents swap relevance labels. By presenting the same document pair under complementary instructions that invert their relevance labels, the training signal forces the retriever to reconsider the same candidate set through the instruction, rather than relying on fixed topical cues. On a 305M-parameter encoder, our method improves performance on the FollowIR benchmark by 45%, surpassing general-purpose embedding models of comparable or larger scale. Through head-to-head comparisons at matched data budgets, we further show that data diversity and instruction supervision play complementary roles: the former preserves general retrieval quality, while the latter improves instruction sensitivity. These results highlight the value of targeted data synthesis for building retrieval systems that are both broadly capable and instruction-aware.
