LogiCoL: Logically-Informed Contrastive Learning for Set-based Dense Retrieval
Not on the map. Why?
Abstract (arXiv)
While significant progress has been made with dual- and bi-encoder dense retrievers, they often struggle on queries with logical connectives, a use case that is often overlooked yet important in downstream applications. Current dense retrievers struggle with such queries, such that the retrieved results do not respect the logical constraints implied in the queries. To address this challenge, we introduce LogiCoL, a logically-informed contrastive learning objective for dense retrievers. LogiCoL builds upon in-batch supervised contrastive learning, and learns dense retrievers to respect the subset and mutually-exclusive set relation between query results via two sets of soft constraints expressed via t-norm in the learning objective. We evaluate the effectiveness of LogiCoL on the task of entity retrieval, where the model is expected to retrieve a set of entities in Wikipedia that satisfy the implicit logical constraints in the query. We show that models trained with LogiCoL yield improvement both in terms of retrieval performance and logical consistency in the results. We provide detailed analysis and insights to uncover why queries with logical connectives are challenging for dense retrievers and why LogiCoL is most effective.
Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 25 of its 34 references that are in the map. First posted on arXiv 26 May 2025.
Builds on
25 of 34 references.
References · 25
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
- 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
- 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
- 2021SimCSE: Simple Contrastive Learning of Sentence Embeddingscited by 2,173
- 2016MS MARCO: A Human Generated MAchine Reading COmprehension Datasetcited by 1,605
- 2022Text Embeddings by Weakly-Supervised Contrastive Pre-trainingcited by 1,253
- 2021Unsupervised Dense Information Retrieval with Contrastive Learningcited by 1,216
- 2020Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrievalcited by 849
- 2017Reading Wikipedia to Answer Open-Domain Questionscited by 777
- 2023Towards General Text Embeddings with Multi-stage Contrastive Learningcited by 723
- 2020Contrastive Learning with Hard Negative Samplescited by 628
- 2020KILT: a Benchmark for Knowledge Intensive Language Taskscited by 573
13 more
- 2021Large Dual Encoders Are Generalizable Retrieverscited by 456
- 2021Efficiently Teaching an Effective Dense Retriever with Balanced Topic Aware Samplingcited by 238
- 2019A Logic-Driven Framework for Consistency of Neural Modelscited by 64
- 2020Joint Constrained Learning for Event-Event Relation Extractioncited by 53
- 2019Augmenting Neural Networks with First-order Logiccited by 51
- 2023QUEST: A Retrieval Dataset of Entity-Seeking Queries with Implicit Set Operationscited by 31
- 2023NevIR: Negation in Neural Information Retrievalcited by 29
- 2024MoDE: CLIP Data Experts via Clusteringcited by 22
- 2023Chain-of-Factors Paper-Reviewer Matchingcited by 11
- 2024BoolQuestions: Does Dense Retrieval Understand Boolean Logic in Language?cited by 6
- 2022Global Contrastive Batch Sampling via Optimization on Sample Permutationscited by 5
- 2024SetBERT: Enhancing Retrieval Performance for Boolean Logic and Set Operation Queriescited by 5
- 2025Constructing Set-Compositional and Negated Representations for First-Stage Rankingcited by 5
Built on by
Citing papers per year: 2026: 3.
3 papers.