Constrained Auto-Regressive Decoding Constrains Generative Retrieval

Abstract

Generative retrieval seeks to replace traditional search index data structures with a single large-scale neural network, offering the potential for improved efficiency and seamless integration with generative large language models. As an end-to-end paradigm, generative retrieval adopts a learned differentiable search index to conduct retrieval by directly generating document identifiers through corpus-specific constrained decoding. The generalization capabilities of generative retrieval on out-of-distribution corpora have gathered significant attention. In this paper, we examine the inherent limitations of constrained auto-regressive generation from two essential perspectives: constraints and beam search. We begin with the Bayes-optimal setting where the generative retrieval model exactly captures the underlying relevance distribution of all possible documents. Then we apply the model to specific corpora by simply adding corpus-specific constraints. Our main findings are two-fold: (i) For the effect of constraints, we derive a lower bound of the error, in terms of the KL divergence between the ground-truth and the model-predicted step-wise marginal distributions. (ii) For the beam search algorithm used during generation, we reveal that the usage of marginal distributions may not be an ideal approach. This paper aims to improve our theoretical understanding of the generalization capabilities of the auto-regressive decoding retrieval paradigm, laying a foundation for its limitations and inspiring future advancements toward more robust and generalizable generative retrieval.

13 pages, 6 figures, 2 tables, accepted by SIGIR 2025 (Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval)

References in corpus (12)

• Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
• Learning Tree-based Deep Model for Recommender Systems
• Rethinking Search: Making Domain Experts out of Dilettantes
• Autoregressive Search Engines: Generating Substrings as Document Identifiers
• CorpusBrain: Pre-train a Generative Retrieval Model for Knowledge-Intensive Language Tasks
• Continual Learning for Generative Retrieval over Dynamic Corpora
• Joint Optimization of Tree-based Index and Deep Model for Recommender Systems
• SLIM: Sparsified Late Interaction for Multi-Vector Retrieval with Inverted Indexes
• From Matching to Generation: A Survey on Generative Information Retrieval
• On the Robustness of Generative Retrieval Models: An Out-of-Distribution Perspective
• Generative Retrieval as Dense Retrieval
• Generative Retrieval with Few-shot Indexing