Setting the Record Straight on Transformer Oversmoothing

• URL: http://arxiv.org/abs/2401.04301v3
• Date: Sat, 27 Jul 2024 07:28:03 GMT
• Title: Setting the Record Straight on Transformer Oversmoothing
• Authors: Gbètondji J-S Dovonon, Michael M. Bronstein, Matt J. Kusner,
• Abstract summary: As model depth increases, Transformers oversmooth, i.e., inputs become more and more similar. We show that smoothing behavior depends on the eigenspectrum of the value and projection weights. Our analysis reveals a simple way to parameterize the weights of the Transformer update equations to influence smoothing behavior.
• Score: 35.125957267464756
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Transformer-based models have recently become wildly successful across a diverse set of domains. At the same time, recent work has shown empirically and theoretically that Transformers are inherently limited. Specifically, they argue that as model depth increases, Transformers oversmooth, i.e., inputs become more and more similar. A natural question is: How can Transformers achieve these successes given this shortcoming? In this work we test these observations empirically and theoretically and uncover a number of surprising findings. We find that there are cases where feature similarity increases but, contrary to prior results, this is not inevitable, even for existing pre-trained models. Theoretically, we show that smoothing behavior depends on the eigenspectrum of the value and projection weights. We verify this empirically and observe that the sign of layer normalization weights can influence this effect. Our analysis reveals a simple way to parameterize the weights of the Transformer update equations to influence smoothing behavior. We hope that our findings give ML researchers and practitioners additional insight into how to develop future Transformer-based models.

Related papers

• Can Looped Transformers Learn to Implement Multi-step Gradient Descent for In-context Learning? [69.4145579827826]
  We show a fast flow on the regression loss despite the gradient non-ity algorithms for our convergence landscape. This is the first theoretical analysis for multi-layer Transformer in this setting.
  arXiv  Detail & Related papers  (2024-10-10T18:29:05Z)
• 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)
• Local to Global: Learning Dynamics and Effect of Initialization for Transformers [20.02103237675619]
  We focus on first-order Markov chains and single-layer transformers. We prove that transformer parameters trained on next-token prediction loss can either converge to global or local minima.
  arXiv  Detail & Related papers  (2024-06-05T08:57:41Z)
• Transformers can optimally learn regression mixture models [22.85684729248361]
  We show that transformers can learn an optimal predictor for mixtures of regressions. Experiments also demonstrate that transformers can learn mixtures of regressions in a sample-efficient fashion. We prove constructively that the decision-theoretic optimal procedure is indeed implementable by a transformer.
  arXiv  Detail & Related papers  (2023-11-14T18:09:15Z)
• Emergent Agentic Transformer from Chain of Hindsight Experience [96.56164427726203]
  We show that a simple transformer-based model performs competitively with both temporal-difference and imitation-learning-based approaches. This is the first time that a simple transformer-based model performs competitively with both temporal-difference and imitation-learning-based approaches.
  arXiv  Detail & Related papers  (2023-05-26T00:43:02Z)
• Transformers learn in-context by gradient descent [58.24152335931036]
  Training Transformers on auto-regressive objectives is closely related to gradient-based meta-learning formulations. We show how trained Transformers become mesa-optimizers i.e. learn models by gradient descent in their forward pass.
  arXiv  Detail & Related papers  (2022-12-15T09:21:21Z)
• Scaling Laws vs Model Architectures: How does Inductive Bias Influence Scaling? [91.78878523252897]
  This paper conducts a systematic study of scaling behaviour of ten diverse model architectures. We show that architecture is an important consideration when performing scaling and the best performing model can fluctuate at different scales.
  arXiv  Detail & Related papers  (2022-07-21T15:50:22Z)
• Transformers from an Optimization Perspective [24.78739299952529]
  We study the problem of finding an energy function underlying the Transformer model. By finding such a function, we can reinterpret Transformers as the unfolding of an interpretable optimization process. This work contributes to our intuition and understanding of Transformers, while potentially laying the ground-work for new model designs.
  arXiv  Detail & Related papers  (2022-05-27T10:45:15Z)

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.