# Rethinking Supervised Pre-training for Better Downstream Transferring

arXiv: https://arxiv.org/abs/2110.06014 · HTML: https://mlsanity.com/paper/2110.06014/

## Key facts

- Cited by 35 papers in the map (data through 14 Oct 2026).
- Builds on 43 of its 65 references that are in the map.
- First posted on arXiv 12 Oct 2021.

## Builds on (most cited first, 30 of 43)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks](https://mlsanity.com/paper/1506.01497/) (2015)
- [Momentum Contrast for Unsupervised Visual Representation Learning](https://mlsanity.com/paper/1911.05722/) (2019)
- [Mask R-CNN](https://mlsanity.com/paper/1703.06870/) (2017)
- [Bootstrap your own latent: A new approach to self-supervised Learning](https://mlsanity.com/paper/2006.07733/) (2020)
- [Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset](https://mlsanity.com/paper/1705.07750/) (2017)
- [Exploring Simple Siamese Representation Learning](https://mlsanity.com/paper/2011.10566/) (2020)
- [Unsupervised Learning of Visual Features by Contrasting Cluster Assignments](https://mlsanity.com/paper/2006.09882/) (2020)
- [Supervised Contrastive Learning](https://mlsanity.com/paper/2004.11362/) (2020)
- [BEiT: BERT Pre-Training of Image Transformers](https://mlsanity.com/paper/2106.08254/) (2021)
- [Improved Baselines with Momentum Contrastive Learning](https://mlsanity.com/paper/2003.04297/) (2020)
- [Describing Textures in the Wild](https://mlsanity.com/paper/1311.3618/) (2013)
- [Unsupervised Domain Adaptation by Backpropagation](https://mlsanity.com/paper/1409.7495/) (2014)
- [Unsupervised Feature Learning via Non-Parametric Instance-level Discrimination](https://mlsanity.com/paper/1805.01978/) (2018)
- [Barlow Twins: Self-Supervised Learning via Redundancy Reduction](https://mlsanity.com/paper/2103.03230/) (2021)
- [Fine-Grained Visual Classification of Aircraft](https://mlsanity.com/paper/1306.5151/) (2013)
- [Unsupervised Representation Learning by Predicting Image Rotations](https://mlsanity.com/paper/1803.07728/) (2018)
- [EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification](https://mlsanity.com/paper/1709.00029/) (2017)
- [Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles](https://mlsanity.com/paper/1603.09246/) (2016)
- [How transferable are features in deep neural networks?](https://mlsanity.com/paper/1411.1792/) (2014)
- [Deep Clustering for Unsupervised Learning of Visual Features](https://mlsanity.com/paper/1807.05520/) (2018)
- [Revisiting Unreasonable Effectiveness of Data in Deep Learning Era](https://mlsanity.com/paper/1707.02968/) (2017)
- [Big Transfer (BiT): General Visual Representation Learning](https://mlsanity.com/paper/1912.11370/) (2019)

## Built on by (most cited first, 30 of 35)

- [SG-Bench: Evaluating LLM Safety Generalization Across Diverse Tasks and Prompt Types](https://mlsanity.com/paper/2410.21965/) (2024)
- [RLIP: Relational Language-Image Pre-training for Human-Object Interaction Detection](https://mlsanity.com/paper/2209.01814/) (2022)
- [AdaNPC: Exploring Non-Parametric Classifier for Test-Time Adaptation](https://mlsanity.com/paper/2304.12566/) (2023)
- [Joint Embedding vs Reconstruction: Provable Benefits of Latent Space Prediction for Self Supervised Learning](https://mlsanity.com/paper/2505.12477/) (2025)
- [Grow and Merge: A Unified Framework for Continuous Categories Discovery](https://mlsanity.com/paper/2210.04174/) (2022)
- [Understanding and Improving Transfer Learning of Deep Models via Neural Collapse](https://mlsanity.com/paper/2212.12206/) (2022)
- [No Reason for No Supervision: Improved Generalization in Supervised Models](https://mlsanity.com/paper/2206.15369/) (2022)
- [Supervised Masked Knowledge Distillation for Few-Shot Transformers](https://mlsanity.com/paper/2303.15466/) (2023)
- [Strong but simple: A Baseline for Domain Generalized Dense Perception by CLIP-based Transfer Learning](https://mlsanity.com/paper/2312.02021/) (2023)
- [MVP: Multi-task Supervised Pre-training for Natural Language Generation](https://mlsanity.com/paper/2206.12131/) (2022)
- [Discriminability-Transferability Trade-Off: An Information-Theoretic Perspective](https://mlsanity.com/paper/2203.03871/) (2022)
- [Deep transfer learning for image classification: a survey](https://mlsanity.com/paper/2205.09904/) (2022)
- [A New Benchmark: On the Utility of Synthetic Data with Blender for Bare Supervised Learning and Downstream Domain Adaptation](https://mlsanity.com/paper/2303.09165/) (2023)
- [Pre-Trained Language-Meaning Models for Multilingual Parsing and Generation](https://mlsanity.com/paper/2306.00124/) (2023)
- [Watch the Neighbors: A Unified K-Nearest Neighbor Contrastive Learning Framework for OOD Intent Discovery](https://mlsanity.com/paper/2210.08909/) (2022)
- [Quantifying the Variability Collapse of Neural Networks](https://mlsanity.com/paper/2306.03440/) (2023)
- [AgriFM: A Multi-source Temporal Remote Sensing Foundation Model for Agriculture Mapping](https://mlsanity.com/paper/2505.21357/) (2025)
- [Revisiting the Encoding of Satellite Image Time Series](https://mlsanity.com/paper/2305.02086/) (2023)
- [Decoupling Pseudo Label Disambiguation and Representation Learning for Generalized Intent Discovery](https://mlsanity.com/paper/2305.17699/) (2023)
- [From Laboratory to Real World: A New Benchmark Towards Privacy-Preserved Visible-Infrared Person Re-Identification](https://mlsanity.com/paper/2503.12232/) (2025)
- [A Survey of Foundation Models for Music Understanding](https://mlsanity.com/paper/2409.09601/) (2024)
- [MaxSup: Overcoming Representation Collapse in Label Smoothing](https://mlsanity.com/paper/2502.15798/) (2025)
- [OPERA: Omni-Supervised Representation Learning with Hierarchical Supervisions](https://mlsanity.com/paper/2210.05557/) (2022)
- [Tackling Data Bias in Painting Classification with Style Transfer](https://mlsanity.com/paper/2301.02524/) (2023)
- [SINCERE: Supervised Information Noise-Contrastive Estimation REvisited](https://mlsanity.com/paper/2309.14277/) (2023)
- [Tree-Regularized Tabular Embeddings](https://mlsanity.com/paper/2403.00963/) (2024)
- [DualFed: Enjoying both Generalization and Personalization in Federated Learning via Hierachical Representations](https://mlsanity.com/paper/2407.17754/) (2024)
- [A Theoretical Framework for Preventing Class Collapse in Supervised Contrastive Learning](https://mlsanity.com/paper/2503.08203/) (2025)
- [A Reality Check of Vision-Language Pre-training in Radiology: Have We Progressed Using Text?](https://mlsanity.com/paper/2504.05227/) (2025)
- [Artificial Intelligence for Geometry-Based Feature Extraction, Analysis and Synthesis in Artistic Images: A Survey](https://mlsanity.com/paper/2412.01450/) (2024)

## Abstract (arXiv)

The pretrain-finetune paradigm has shown outstanding performance on many applications of deep learning, where a model is pre-trained on a upstream large dataset (e.g. ImageNet), and is then fine-tuned to different downstream tasks. Though for most cases, the pre-training stage is conducted based on supervised methods, recent works on self-supervised pre-training have shown powerful transferability and even outperform supervised pre-training on multiple downstream tasks. It thus remains an open question how to better generalize supervised pre-training model to downstream tasks. In this paper, we argue that the worse transferability of existing supervised pre-training methods arise from the negligence of valuable intra-class semantic difference. This is because these methods tend to push images from the same class close to each other despite of the large diversity in their visual contents, a problem to which referred as "overfit of upstream tasks". To alleviate this problem, we propose a new supervised pre-training method based on Leave-One-Out K-Nearest-Neighbor, or LOOK for short. It relieves the problem of overfitting upstream tasks by only requiring each image to share its class label with most of its k nearest neighbors, thus allowing each class to exhibit a multi-mode distribution and consequentially preserving part of intra-class difference for better transferring to downstream tasks. We developed efficient implementation of the proposed method that scales well to large datasets. Experimental studies on multiple downstream tasks show that LOOK outperforms other state-of-the-art methods for supervised and self-supervised pre-training.