Understanding Generalization in Transformers: Error Bounds and Training Dynamics Under Benign and Harmful Overfitting

• URL: http://arxiv.org/abs/2502.12508v1
• Date: Tue, 18 Feb 2025 03:46:01 GMT
• Title: Understanding Generalization in Transformers: Error Bounds and Training Dynamics Under Benign and Harmful Overfitting
• Authors: Yingying Zhang, Zhenyu Wu, Jian Li, Yong Liu,
• Abstract summary: We develop a generalization theory for a two-layer transformer with labeled flip noise. We present generalization error bounds for both benign and harmful overfitting under varying signal-to-noise ratios. We conduct extensive experiments to identify key factors that influence test errors in transformers.
• Score: 36.149708427591534
• License:
• Abstract: Transformers serve as the foundational architecture for many successful large-scale models, demonstrating the ability to overfit the training data while maintaining strong generalization on unseen data, a phenomenon known as benign overfitting. However, research on how the training dynamics influence error bounds within the context of benign overfitting has been limited. This paper addresses this gap by developing a generalization theory for a two-layer transformer with labeled flip noise. Specifically, we present generalization error bounds for both benign and harmful overfitting under varying signal-to-noise ratios (SNR), where the training dynamics are categorized into three distinct stages, each with its corresponding error bounds. Additionally, we conduct extensive experiments to identify key factors that influence test errors in transformers. Our experimental results align closely with the theoretical predictions, validating our findings.

Related papers

• Interpreting Affine Recurrence Learning in GPT-style Transformers [54.01174470722201]
  In-context learning allows GPT-style transformers to generalize during inference without modifying their weights. This paper focuses specifically on their ability to learn and predict affine recurrences as an ICL task. We analyze the model's internal operations using both empirical and theoretical approaches.
  arXiv  Detail & Related papers  (2024-10-22T21:30:01Z)
• On the Training Convergence of Transformers for In-Context Classification of Gaussian Mixtures [20.980349268151546]
  This work aims to theoretically study the training dynamics of transformers for in-context classification tasks. We demonstrate that, for in-context classification of Gaussian mixtures under certain assumptions, a single-layer transformer trained via gradient descent converges to a globally optimal model at a linear rate.
  arXiv  Detail & Related papers  (2024-10-15T16:57:14Z)
• Unveil Benign Overfitting for Transformer in Vision: Training Dynamics, Convergence, and Generalization [88.5582111768376]
  We study the optimization of a Transformer composed of a self-attention layer with softmax followed by a fully connected layer under gradient descent on a certain data distribution model. Our results establish a sharp condition that can distinguish between the small test error phase and the large test error regime, based on the signal-to-noise ratio in the data model.
  arXiv  Detail & Related papers  (2024-09-28T13:24:11Z)
• Learning on Transformers is Provable Low-Rank and Sparse: A One-layer Analysis [63.66763657191476]
  We show that efficient numerical training and inference algorithms as low-rank computation have impressive performance for learning Transformer-based adaption. We analyze how magnitude-based models affect generalization while improving adaption. We conclude that proper magnitude-based has a slight on the testing performance.
  arXiv  Detail & Related papers  (2024-06-24T23:00:58Z)
• On robust overfitting: adversarial training induced distribution matters [32.501773057885735]
  Adversarial training may be regarded as standard training with a modified loss function. But its generalization error appears much larger than standard training under standard loss. This phenomenon, known as robust overfitting, has attracted significant research attention and remains largely as a mystery.
  arXiv  Detail & Related papers  (2023-11-28T05:11:53Z)
• In-Context Convergence of Transformers [63.04956160537308]
  We study the learning dynamics of a one-layer transformer with softmax attention trained via gradient descent. For data with imbalanced features, we show that the learning dynamics take a stage-wise convergence process.
  arXiv  Detail & Related papers  (2023-10-08T17:55:33Z)
• The Role of Mutual Information in Variational Classifiers [47.10478919049443]
  We study the generalization error of classifiers relying on encodings trained on the cross-entropy loss. We derive bounds to the generalization error showing that there exists a regime where the generalization error is bounded by the mutual information.
  arXiv  Detail & Related papers  (2020-10-22T12:27:57Z)
• Learning perturbation sets for robust machine learning [97.6757418136662]
  We use a conditional generator that defines the perturbation set over a constrained region of the latent space. We measure the quality of our learned perturbation sets both quantitatively and qualitatively. We leverage our learned perturbation sets to train models which are empirically and certifiably robust to adversarial image corruptions and adversarial lighting variations.
  arXiv  Detail & Related papers  (2020-07-16T16:39:54Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.