The paper introduces Natural Hard-Negative Mining (NHNM) for instruction‑following information retrieval, a method that turns original positive passages into hard negatives under query‑compatible instructions. NHNM uses a pipeline of Candidate Cleaning, Reversal Construction, and Geometry Filtering to generate a compact, high‑contrast training set called NHNM‑MARCO with nearly 7,000 instances. Models fine‑tuned on NHNM‑MARCO, ranging from 568 M to 4 B parameters, achieve strong instruction‑following performance on InstructIR and FollowIR, matching or exceeding larger retrievers and demonstrating that the gains stem from natural hard‑negative supervision rather than model size or data volume.
arXiv:2606. 01304v1 Announce Type: new Abstract: Hard negative mining has become the dominant strategy for training retrievers, yet it faces intrinsic limitations: negatives are bounded by corpus availability, selected by retriever score rather than diagnostic value, and increasingly contaminated by false positives as the retriever improves.
arXiv:2610.35045v1 Announce Type: cross Abstract: Dense retrievers are commonly trained with hard negatives selected from lexical or dense retrieval results. Retrieval proximity, however, does not in...
Task-Level Natural Language Priors as Learning Signals for Low-Resource LLM Training proposes Prior-Guided Tuning (PGT), a training approach that treats natural-language priors as auxiliary learning signals rather than just input context. The method introduces Contrastive Prior Steering (CPS), which adds positive and negative prior-conditioned auxiliary losses while preserving the original supervised objective. Experiments on AmbiMath, Jigsaw, and MNLI/HANS demonstrate that CPS consistently outperforms plain and prompt fine-tuning, achieving high accuracy and significant gains with limited training data.
The study investigates which properties of hard‑negative supervision predict dense‑retriever transfer before training. It compares lexical non‑coverage, frozen‑encoder similarity, hardness proxies, and their combinations across BM25, dense, LLM‑generated, and hybrid sources for MS MARCO training. The results show that mean lexical residuality, an IDF‑weighted measure of query‑token non‑coverage, consistently reproduces the aggregate transfer ordering, while semantic statistics alone or combined with lexical signals do not improve rank agreement.
arXiv:2603. 20990v2 Announce Type: replace-cross Abstract: Hard-negative source selection for dense retrieval is usually decided only after fine-tuning and downstream evaluation.
arXiv:2608.29604v1 Announce Type: cross Abstract: Vision-language retrieval with CLIP-style dual encoders achieves strong cross-modal performance, yet practical accuracy often hinges on localized sem...
arXiv:2605. 27023v2 Announce Type: replace Abstract: Knowledge graphs (KGs) have become the core backbone of numerous downstream tasks such as question answering and recommender systems.
arXiv:2603. 20990v3 Announce Type: replace-cross Abstract: Hard-negative source selection for dense retrieval is usually decided only after fine-tuning and downstream evaluation.
E‑SENS is a training‑free reranking technique designed to improve negative‑constraint retrieval in retrieval‑augmented language models. It works by extracting a compact ‘trap query’ that represents the excluded concept and subtracting the similarity of this trap query from the original retrieval score, thereby reducing the influence of documents that mention the excluded concept. Experiments on the ExcluIR benchmark demonstrate that E‑SENS achieves a clear recall‑violation trade‑off across four embedding models and effectively reduces trap retrieval while preserving recall.
arXiv:2610.09700v1 Announce Type: new Abstract: Synthetic hard negatives generated in the representation space have proven effective for unimodal self-supervised learning, but transferring this idea...
arXiv:2508. 00955v3 Announce Type: replace-cross Abstract: Adapting generative Multimodal Large Language Models (MLLMs) into universal embedding models typically demands resource-intensive contrastive pre-training, while traditional hard negative mining methods suffer from severe false negative contamination.
The paper introduces DEPT, a method that trains a single decoder-only large language model to both expand queries and encode documents for retrieval. By preserving document embeddings close to their initial cached values while allowing gradients to flow through the generator, DEPT stabilizes retrieval targets and enables efficient index reuse and online hard‑negative mining. Experiments on the BEIR benchmark with Qwen3‑4B‑Instruct‑2507 and LLaMA‑3.2‑3B‑Instruct show that DEPT outperforms training‑free, independently trained, and staged unified baselines, with ablations confirming the benefits of preservation, whitening, end‑to‑end expansion training, and online negatives.