# The Advantage of Fine-Grained Training

arXiv: https://arxiv.org/abs/2509.05130 · HTML: https://mlsanity.com/paper/2509.05130/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 50 references that are in the map.
- First posted on arXiv 5 Sep 2025.

## Builds on (most cited first, 27 of 27)

- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms](https://mlsanity.com/paper/1708.07747/) (2017)
- [On the Opportunities and Risks of Foundation Models](https://mlsanity.com/paper/2108.07258/) (2021)
- [Shortcut Learning in Deep Neural Networks](https://mlsanity.com/paper/2004.07780/) (2020)
- [How transferable are features in deep neural networks?](https://mlsanity.com/paper/1411.1792/) (2014)
- [Deep Learning and the Information Bottleneck Principle](https://mlsanity.com/paper/1503.02406/) (2015)
- [Taskonomy: Disentangling Task Transfer Learning](https://mlsanity.com/paper/1804.08328/) (2018)
- [ImageNet-21K Pretraining for the Masses](https://mlsanity.com/paper/2104.10972/) (2021)
- [Do Better ImageNet Models Transfer Better?](https://mlsanity.com/paper/1805.08974/) (2018)
- [Deep Learning for Classical Japanese Literature](https://mlsanity.com/paper/1812.01718/) (2018)
- [Multi-Task Learning with Deep Neural Networks: A Survey](https://mlsanity.com/paper/2009.09796/) (2020)
- [Emergence of Invariance and Disentanglement in Deep Representations](https://mlsanity.com/paper/1706.01350/) (2017)
- [No Subclass Left Behind: Fine-Grained Robustness in Coarse-Grained Classification Problems](https://mlsanity.com/paper/2011.12945/) (2020)
- [The Power of Depth for Feedforward Neural Networks](https://mlsanity.com/paper/1512.03965/) (2015)
- [Why does deep and cheap learning work so well?](https://mlsanity.com/paper/1608.08225/) (2016)
- [Theoretical Issues in Deep Networks: Approximation, Optimization and Generalization](https://mlsanity.com/paper/1908.09375/) (2019)
- [How Deep Neural Networks Learn Compositional Data: The Random Hierarchy Model](https://mlsanity.com/paper/2307.02129/) (2023)
- [The Information Complexity of Learning Tasks, their Structure and their Distance](https://mlsanity.com/paper/1904.03292/) (2019)
- [A hierarchical loss and its problems when classifying non-hierarchically](https://mlsanity.com/paper/1709.01062/) (2017)
- [On the Trade-off of Intra-/Inter-class Diversity for Supervised Pre-training](https://mlsanity.com/paper/2305.12224/) (2023)
- [Position: A Theory of Deep Learning Must Include Compositional Sparsity](https://mlsanity.com/paper/2507.02550/) (2025)
- [Understanding the Impact of Label Granularity on CNN-based Image Classification](https://mlsanity.com/paper/1901.07012/) (2019)
- [Learn Class Hierarchy using Convolutional Neural Networks](https://mlsanity.com/paper/2005.08622/) (2020)
- [A Simple and Effective Framework for Strict Zero-Shot Hierarchical Classification](https://mlsanity.com/paper/2305.15282/) (2023)
- [Intraclass clustering: an implicit learning ability that regularizes DNNs](https://mlsanity.com/paper/2103.06733/) (2021)
- [Inversion dynamics of class manifolds in deep learning reveals tradeoffs underlying generalisation](https://mlsanity.com/paper/2303.05161/) (2023)
- [Why Fine-grained Labels in Pretraining Benefit Generalization?](https://mlsanity.com/paper/2410.23129/) (2024)

## Built on by (most cited first, 1 of 1)

- [Extremely coarse learning objectives induce human-aligned representations in AI vision models](https://mlsanity.com/paper/2605.05556/) (2026)

## Abstract (arXiv)

In classification problems, models are trained to predict a class label based on the input data features. However, class labels are organized hierarchically in many datasets. While a classification task is often defined at a specific level of this hierarchy, training can utilize a finer granularity of labels. Empirical evidence suggests that such fine-grained training can enhance performance. In this work, we investigate the generality of this observation and explore its underlying causes using both real and synthetic datasets. We show that training on fine-grained labels does not universally improve classification accuracy. Instead, the effectiveness of this strategy depends on the geometric structure of the data and its relations with the label hierarchy. Specifically, we show that the advantage of fine-grained training crucially depends on the degree of alignment between the decision boundaries required for the fine- and coarse-grained tasks, a property that we term boundary redundancy. Additionally, factors such as dataset size and model capacity significantly influence whether fine-grained labels provide a performance benefit. Indeed, we identify a transition, whose location is largely controlled by the degree of overparameterization, separating regimes where fine-grained training improves performance from those where direct coarse-grained training is preferable.