Learning Effective Representations for Retrieval Using Self-Distillation with Adaptive Relevance Margins
원문 초록
Representation-based retrieval models, so-called biencoders, estimate the relevance of a document to a query by calculating the similarity of their respective embeddings. Current state-of-the-art biencoders are trained using an expensive training regime involving knowledge distillation from a teacher model and batch-sampling. Instead of relying on a teacher model, we contribute a novel parameter-free loss function for self-supervision that exploits the pre-trained language modeling capabilities of the encoder model as a training signal, eliminating the need for batch sampling by performing implicit hard negative mining. We investigate the capabilities of our proposed approach through extensive ablation studies, demonstrating that self-distillation can match the effectiveness of teacher distillation using only 13.5% of the data, while offering a speedup in training time between 3x and 15x compared to parametrized losses. Code and data is made openly available.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Kodezi Chronos: A Debugging-First Language Model for Repository-Scale, Memory-Driven Code Understanding
Differential Mamba
RMIT-ADM+S at the SIGIR 2025 LiveRAG Challenge
Visual-Language Model Knowledge Distillation Method for Image Quality Assessment
Llama Nemoretriever Colembed: Top-Performing Text-Image Retrieval Model
Case-based Reasoning Augmented Large Language Model Framework for Decision Making in Realistic Safety-Critical Driving Scenarios
토론
이 논문과 연결된 포럼 글이 아직 없습니다.