Lightweight And Direct Document Relevance Optimization For Generative Information Retrieval

Abstract

Generative information retrieval (GenIR) is a promising neural retrieval paradigm that formulates document retrieval as a document identifier (docid) generation task, allowing for end-to-end optimization toward a unified global retrieval objective. However, existing GenIR models suffer from token-level misalignment, where models trained to predict the next token often fail to capture document-level relevance effectively. While reinforcement learning-based methods, such as reinforcement learning from relevance feedback (RLRF), aim to address this misalignment through reward modeling, they introduce significant complexity, requiring the optimization of an auxiliary reward function followed by reinforcement fine-tuning, which is computationally expensive and often unstable. To address these challenges, we propose direct document relevance optimization (DDRO), which aligns token-level docid generation with document-level relevance estimation through direct optimization via pairwise ranking

Stats

• citations3
• S2 citations—
• github stars0
• HF likes0
• heat score4.52
• arxiv keymekonnen2025lightweight

Related papers

• Generative Retrieval Meets Multi-graded Relevance (2024)2.26
• Learning To Tokenize For Generative Retrieval (2023)4.52
• Scalable And Effective Generative Information Retrieval (2023)10.48
• Does Generative Retrieval Overcome The Limitations Of Dense Retrieval? (2025)0.00
• Listwise Generative Retrieval Models Via A Sequential Learning Process (2024)8.60
• Generative Retrieval As Multi-vector Dense Retrieval (2024)8.60
• Differentiable Geometric Indexing For End-to-end Generative Retrieval (2026)0.00
• Irgen: Generative Modeling For Image Retrieval (2023)7.16