Deep Fusion: Capturing Dependencies in Contrastive Learning via Transformer Projection Heads
Fuente:
arXiv
Salvato in:
| Autori principali: | , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909337953239040 |
|---|---|
| author | Li, Huanran Pimentel-Alarcón, Daniel |
| author_facet | Li, Huanran Pimentel-Alarcón, Daniel |
| contents | Contrastive Learning (CL) has emerged as a powerful method for training feature extraction models using unlabeled data. Recent studies suggest that incorporating a linear projection head post-backbone significantly enhances model performance. In this work, we investigate the use of a transformer model as a projection head within the CL framework, aiming to exploit the transformer's capacity for capturing long-range dependencies across embeddings to further improve performance. Our key contributions are fourfold: First, we introduce a novel application of transformers in the projection head role for contrastive learning, marking the first endeavor of its kind. Second, our experiments reveal a compelling "Deep Fusion" phenomenon where the attention mechanism progressively captures the correct relational dependencies among samples from the same class in deeper layers. Third, we provide a theoretical framework that explains and supports this "Deep Fusion" behavior. Finally, we demonstrate through experimental results that our model achieves superior performance compared to the existing approach of using a feed-forward layer. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_18681 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Deep Fusion: Capturing Dependencies in Contrastive Learning via Transformer Projection Heads Li, Huanran Pimentel-Alarcón, Daniel Machine Learning Artificial Intelligence Contrastive Learning (CL) has emerged as a powerful method for training feature extraction models using unlabeled data. Recent studies suggest that incorporating a linear projection head post-backbone significantly enhances model performance. In this work, we investigate the use of a transformer model as a projection head within the CL framework, aiming to exploit the transformer's capacity for capturing long-range dependencies across embeddings to further improve performance. Our key contributions are fourfold: First, we introduce a novel application of transformers in the projection head role for contrastive learning, marking the first endeavor of its kind. Second, our experiments reveal a compelling "Deep Fusion" phenomenon where the attention mechanism progressively captures the correct relational dependencies among samples from the same class in deeper layers. Third, we provide a theoretical framework that explains and supports this "Deep Fusion" behavior. Finally, we demonstrate through experimental results that our model achieves superior performance compared to the existing approach of using a feed-forward layer. |
| title | Deep Fusion: Capturing Dependencies in Contrastive Learning via Transformer Projection Heads |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2403.18681 |