Transformers for Learning on Noisy and Task-Level Manifolds: Approximation and Generalization Insights
It provides foundational theoretical insights for transformers in noisy, structured data settings, which is incremental but addresses a gap in understanding for models like GPT and SORA.
This paper tackles the problem of transformers learning from noisy data on low-dimensional manifolds by establishing theoretical bounds on approximation and generalization errors, showing that performance depends on the intrinsic dimension of the manifold rather than the noise.
Transformers serve as the foundational architecture for large language and video generation models, such as GPT, BERT, SORA and their successors. Empirical studies have demonstrated that real-world data and learning tasks exhibit low-dimensional structures, along with some noise or measurement error. The performance of transformers tends to depend on the intrinsic dimension of the data/tasks, though theoretical understandings remain largely unexplored for transformers. This work establishes a theoretical foundation by analyzing the performance of transformers for regression tasks involving noisy input data on a manifold. Specifically, the input data are in a tubular neighborhood of a manifold, while the ground truth function depends on the projection of the noisy data onto the manifold. We prove approximation and generalization errors which crucially depend on the intrinsic dimension of the manifold. Our results demonstrate that transformers can leverage low-complexity structures in learning task even when the input data are perturbed by high-dimensional noise. Our novel proof technique constructs representations of basic arithmetic operations by transformers, which may hold independent interest.