Initialization is Critical to Whether Transformers Fit Composite Functions by Reasoning or Memorizing

• URL: http://arxiv.org/abs/2405.05409v5
• Date: Mon, 13 Jan 2025 11:35:37 GMT
• Title: Initialization is Critical to Whether Transformers Fit Composite Functions by Reasoning or Memorizing
• Authors: Zhongwang Zhang, Pengxiao Lin, Zhiwei Wang, Yaoyu Zhang, Zhi-Qin John Xu,
• Abstract summary: Transformers have shown impressive capabilities across various tasks, but their performance on compositional problems remains a topic of debate. We discover that the parameter initialization scale plays a critical role in determining whether the model learns inferential (reasoning-based) solutions. We further find that inferential (reasoning-based) solutions exhibit low complexity bias, which we hypothesize is a key factor enabling them to learn individual mappings for single anchors.
• Score: 10.206921909332006
• License:
• Abstract: Transformers have shown impressive capabilities across various tasks, but their performance on compositional problems remains a topic of debate. In this work, we investigate the mechanisms of how transformers behave on unseen compositional tasks. We discover that the parameter initialization scale plays a critical role in determining whether the model learns inferential (reasoning-based) solutions, which capture the underlying compositional primitives, or symmetric (memory-based) solutions, which simply memorize mappings without understanding the compositional structure. By analyzing the information flow and vector representations within the model, we reveal the distinct mechanisms underlying these solution types. We further find that inferential (reasoning-based) solutions exhibit low complexity bias, which we hypothesize is a key factor enabling them to learn individual mappings for single anchors. We validate our conclusions on various real-world datasets. Our findings provide valuable insights into the role of initialization scale in tuning the reasoning and memorizing ability and we propose the initialization rate $\gamma$ to be a convenient tunable hyper-parameter in common deep learning frameworks, where $1/d_{\mathrm{in}}^\gamma$ is the standard deviation of parameters of the layer with $d_{\mathrm{in}}$ input neurons.

Related papers

• Complexity Control Facilitates Reasoning-Based Compositional Generalization in Transformers [10.206921909332006]
  This study investigates the internal mechanisms underlying Transformers' behavior in compositional tasks. We find that complexity control strategies influence whether the model learns primitive-level rules that generalize out-of-distribution (reasoning-based solutions) or relies solely on memorized mappings (memory-based solutions)
  arXiv  Detail & Related papers  (2025-01-15T02:54:52Z)
• Counting in Small Transformers: The Delicate Interplay between Attention and Feed-Forward Layers [16.26331213222281]
  We investigate how architectural design choices influence the space of solutions that a transformer can implement and learn. We characterize two different counting strategies that small transformers can implement theoretically. Our findings highlight that even in simple settings, slight variations in model design can cause significant changes to the solutions a transformer learns.
  arXiv  Detail & Related papers  (2024-07-16T09:48:10Z)
• From Bricks to Bridges: Product of Invariances to Enhance Latent Space Communication [19.336940758147442]
  It has been observed that representations learned by distinct neural networks conceal structural similarities when the models are trained under similar inductive biases. We introduce a versatile method to directly incorporate a set of invariances into the representations, constructing a product space of invariant components on top of the latent representations. We validate our solution on classification and reconstruction tasks, observing consistent latent similarity and downstream performance improvements in a zero-shot stitching setting.
  arXiv  Detail & Related papers  (2023-10-02T13:55:38Z)
• DIFFormer: Scalable (Graph) Transformers Induced by Energy Constrained Diffusion [66.21290235237808]
  We introduce an energy constrained diffusion model which encodes a batch of instances from a dataset into evolutionary states. We provide rigorous theory that implies closed-form optimal estimates for the pairwise diffusion strength among arbitrary instance pairs. Experiments highlight the wide applicability of our model as a general-purpose encoder backbone with superior performance in various tasks.
  arXiv  Detail & Related papers  (2023-01-23T15:18:54Z)
• Equivariance with Learned Canonicalization Functions [77.32483958400282]
  We show that learning a small neural network to perform canonicalization is better than using predefineds. Our experiments show that learning the canonicalization function is competitive with existing techniques for learning equivariant functions across many tasks.
  arXiv  Detail & Related papers  (2022-11-11T21:58:15Z)
• Transformer for Partial Differential Equations' Operator Learning [0.0]
  We present an attention-based framework for data-driven operator learning, which we term Operator Transformer (OFormer) Our framework is built upon self-attention, cross-attention, and a set of point-wise multilayer perceptrons (MLPs)
  arXiv  Detail & Related papers  (2022-05-26T23:17:53Z)
• Amortized Inference for Causal Structure Learning [72.84105256353801]
  Learning causal structure poses a search problem that typically involves evaluating structures using a score or independence test. We train a variational inference model to predict the causal structure from observational/interventional data. Our models exhibit robust generalization capabilities under substantial distribution shift.
  arXiv  Detail & Related papers  (2022-05-25T17:37:08Z)
• Generalization of Neural Combinatorial Solvers Through the Lens of Adversarial Robustness [68.97830259849086]
  Most datasets only capture a simpler subproblem and likely suffer from spurious features. We study adversarial robustness - a local generalization property - to reveal hard, model-specific instances and spurious features. Unlike in other applications, where perturbation models are designed around subjective notions of imperceptibility, our perturbation models are efficient and sound. Surprisingly, with such perturbations, a sufficiently expressive neural solver does not suffer from the limitations of the accuracy-robustness trade-off common in supervised learning.
  arXiv  Detail & Related papers  (2021-10-21T07:28:11Z)
• A Trainable Optimal Transport Embedding for Feature Aggregation and its Relationship to Attention [96.77554122595578]
  We introduce a parametrized representation of fixed size, which embeds and then aggregates elements from a given input set according to the optimal transport plan between the set and a trainable reference. Our approach scales to large datasets and allows end-to-end training of the reference, while also providing a simple unsupervised learning mechanism with small computational cost.
  arXiv  Detail & Related papers  (2020-06-22T08:35:58Z)
• Joint learning of variational representations and solvers for inverse problems with partially-observed data [13.984814587222811]
  In this paper, we design an end-to-end framework allowing to learn actual variational frameworks for inverse problems in a supervised setting. The variational cost and the gradient-based solver are both stated as neural networks using automatic differentiation for the latter. This leads to a data-driven discovery of variational models.
  arXiv  Detail & Related papers  (2020-06-05T19:53:34Z)
• Learning What Makes a Difference from Counterfactual Examples and Gradient Supervision [57.14468881854616]
  We propose an auxiliary training objective that improves the generalization capabilities of neural networks. We use pairs of minimally-different examples with different labels, a.k.a counterfactual or contrasting examples, which provide a signal indicative of the underlying causal structure of the task. Models trained with this technique demonstrate improved performance on out-of-distribution test sets.
  arXiv  Detail & Related papers  (2020-04-20T02:47:49Z)

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.