Diminishing Returns in Self-Supervised Learning

📅 2025-12-03
📈 Citations: 0
Influential: 0
📄 PDF

career value

180K/year
🤖 AI Summary
This study investigates the marginal benefits of a three-stage strategy—self-supervised pretraining, intermediate fine-tuning, and downstream task adaptation—for small-scale Vision Transformers (~5M parameters). Motivated by the observation that intermediate fine-tuning may degrade downstream performance due to task misalignment, we propose a systematic ablation framework to assess the impact of varying dataset and objective combinations across stages. Experiments reveal that targeted pretraining substantially improves small-model performance, whereas introducing semantically distant intermediate tasks yields no gain—and often harms performance while wasting compute. The core contribution is the empirical demonstration that, for small ViTs, **the quality of data selection is far more critical than the number of stacked tasks**, challenging conventional assumptions about multi-stage transfer. This finding provides key empirical evidence and methodological guidance for designing efficient, lightweight self-supervised learning paradigms.

Technology Category

Application Category

📝 Abstract
While transformer-based architectures have taken computer vision and NLP by storm, they often require a vast amount of parameters and training data to attain strong performance. In this work, we experiment with three distinct pre-training, intermediate fine-tuning, and downstream datasets and training objectives to explore their marginal benefits on a small 5M-parameter vision transformer. We find that while pre-training and fine-tuning always help our model but have diminishing returns, intermediate fine-tuning can actually show harmful impact on downstream performance, potentially due to dissimilarity in task mechanics. Taken together, our results suggest that small-scale ViTs benefit most from targeted pre-training and careful data selection, while indiscriminate stacking of intermediate tasks can waste compute and even degrade performance.
Problem

Research questions and friction points this paper is trying to address.

Explores diminishing returns in self-supervised learning for small vision transformers.
Investigates how intermediate fine-tuning can harm downstream task performance.
Identifies targeted pre-training and data selection as key for efficient small models.
Innovation

Methods, ideas, or system contributions that make the work stand out.

Targeted pre-training for small-scale vision transformers
Careful data selection to avoid diminishing returns
Avoiding indiscriminate intermediate fine-tuning to prevent performance degradation
O
Oli Bridge
University College London, London UK
H
Huey Sun
University College London, London UK
B
Botond Branyicskai-Nagy
University College London, London UK
C
Charles D'Ornano
University College London, London UK
S
Shomit Basu
University College London, London UK