LGAI-EMBEDDING-Preview Technical Report

TL;DR

This paper introduces a unified instruction-based framework for learning versatile text embeddings using a decoder-only LLM, achieving strong performance across diverse IR and non-IR tasks without task-specific fine-tuning.

Contribution

It proposes a novel approach combining in-context learning, soft supervision, and adaptive hard-negative mining to generate high-quality, generalized text embeddings.

Findings

Achieves top performance on the MTEB benchmark across 41 tasks.

Outperforms larger or fully fine-tuned models in generalization.

Demonstrates effective use of soft relevance scores and adaptive negative sampling.

Abstract

This report presents a unified instruction-based framework for learning generalized text embeddings optimized for both information retrieval (IR) and non-IR tasks. Built upon a decoder-only large language model (Mistral-7B), our approach combines in-context learning, soft supervision, and adaptive hard-negative mining to generate context-aware embeddings without task-specific fine-tuning. Structured instructions and few-shot examples are used to guide the model across diverse tasks, enabling strong performance on classification, semantic similarity, clustering, and reranking benchmarks. To improve semantic discrimination, we employ a soft labeling framework where continuous relevance scores, distilled from a high-performance dense retriever and reranker, serve as fine-grained supervision signals. In addition, we introduce adaptive margin-based hard-negative mining, which filters out semantically ambiguous negatives based on their similarity to positive examples, thereby enhancing training stability and retrieval robustness. Our model is evaluated on the newly introduced MTEB (English, v2) benchmark, covering 41 tasks across seven categories. Results show that our method achieves strong generalization and ranks among the top-performing models by Borda score, outperforming several larger or fully fine-tuned baselines. These findings highlight the effectiveness of combining in-context prompting, soft supervision, and adaptive sampling for scalable, high-quality embedding generation.