# Improving zero-shot learning by mitigating the hubness problem

arXiv: https://arxiv.org/abs/1412.6568 · HTML: https://mlsanity.com/paper/1412.6568/

## Key facts

- Cited by 54 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Dec 2014.

## Built on by (most cited first, 30 of 54)

- [Cross Modal Retrieval with Querybank Normalisation](https://mlsanity.com/paper/2112.12777/) (2021)
- [A Survey on Open-Vocabulary Detection and Segmentation: Past, Present, and Future](https://mlsanity.com/paper/2307.09220/) (2023)
- [Black Box Few-Shot Adaptation for Vision-Language models](https://mlsanity.com/paper/2304.01752/) (2023)
- [Zero-Shot Learning on 3D Point Cloud Objects and Beyond](https://mlsanity.com/paper/2104.04980/) (2021)
- [Primitive Generation and Semantic-related Alignment for Universal Zero-Shot Segmentation](https://mlsanity.com/paper/2306.11087/) (2023)
- [Part-aware Unified Representation of Language and Skeleton for Zero-shot Action Recognition](https://mlsanity.com/paper/2406.13327/) (2024)
- [Balance Act: Mitigating Hubness in Cross-Modal Retrieval with Query and Gallery Banks](https://mlsanity.com/paper/2310.11612/) (2023)
- [Bilingual Lexicon Induction via Unsupervised Bitext Construction and Word Alignment](https://mlsanity.com/paper/2101.00148/) (2021)
- [Relation Matters: Foreground-aware Graph-based Relational Reasoning for Domain Adaptive Object Detection](https://mlsanity.com/paper/2206.02355/) (2022)
- [Beyond English-Centric Bitexts for Better Multilingual Language Representation Learning](https://mlsanity.com/paper/2210.14867/) (2022)
- [Semantic-Promoted Debiasing and Background Disambiguation for Zero-Shot Instance Segmentation](https://mlsanity.com/paper/2305.13173/) (2023)
- [Hierarchical Visual Primitive Experts for Compositional Zero-Shot Learning](https://mlsanity.com/paper/2308.04016/) (2023)
- [Does injecting linguistic structure into language models lead to better alignment with brain recordings?](https://mlsanity.com/paper/2101.12608/) (2021)
- [Combining Static Word Embeddings and Contextual Representations for Bilingual Lexicon Induction](https://mlsanity.com/paper/2106.03084/) (2021)
- [Simple Baselines for Interactive Video Retrieval with Questions and Answers](https://mlsanity.com/paper/2308.10402/) (2023)
- [Clustering with UMAP: Why and How Connectivity Matters](https://mlsanity.com/paper/2108.05525/) (2021)
- [Majority Voting with Bidirectional Pre-translation For Bitext Retrieval](https://mlsanity.com/paper/2103.06369/) (2021)
- [Cross-language Sentence Selection via Data Augmentation and Rationale Training](https://mlsanity.com/paper/2106.02293/) (2021)
- [Multi-Label Generalized Zero Shot Learning for the Classification of Disease in Chest Radiographs](https://mlsanity.com/paper/2107.06563/) (2021)
- [Training Cross-Lingual embeddings for Setswana and Sepedi](https://mlsanity.com/paper/2111.06230/) (2021)
- [Structured Latent Embeddings for Recognizing Unseen Classes in Unseen Domains](https://mlsanity.com/paper/2107.05622/) (2021)
- [Vocabulary-informed Zero-shot and Open-set Learning](https://mlsanity.com/paper/2301.00998/) (2023)
- [Towards Higher-order Topological Consistency for Unsupervised Network Alignment](https://mlsanity.com/paper/2208.12463/) (2022)
- [Align before Search: Aligning Ads Image to Text for Accurate Cross-Modal Sponsored Search](https://mlsanity.com/paper/2309.16141/) (2023)
- [Prediction hubs are context-informed frequent tokens in LLMs](https://mlsanity.com/paper/2502.10201/) (2025)
- [SATTC: Structure-Aware Label-Free Test-Time Calibration for Cross-Subject EEG-to-Image Retrieval](https://mlsanity.com/paper/2603.20738/) (2026)
- [Cross-Lingual Phrase Retrieval](https://mlsanity.com/paper/2204.08887/) (2022)
- [Resolving Semantic Confusions for Improved Zero-Shot Detection](https://mlsanity.com/paper/2212.06097/) (2022)
- [ADELT: Transpilation Between Deep Learning Frameworks](https://mlsanity.com/paper/2303.03593/) (2023)
- [CommonsenseVIS: Visualizing and Understanding Commonsense Reasoning Capabilities of Natural Language Models](https://mlsanity.com/paper/2307.12382/) (2023)

## Abstract (arXiv)

The zero-shot paradigm exploits vector-based word representations extracted from text corpora with unsupervised methods to learn general mapping functions from other feature spaces onto word space, where the words associated to the nearest neighbours of the mapped vectors are used as their linguistic labels. We show that the neighbourhoods of the mapped elements are strongly polluted by hubs, vectors that tend to be near a high proportion of items, pushing their correct labels down the neighbour list. After illustrating the problem empirically, we propose a simple method to correct it by taking the proximity distribution of potential neighbours across many mapped vectors into account. We show that this correction leads to consistent improvements in realistic zero-shot experiments in the cross-lingual, image labeling and image retrieval domains.