Forgetting Transformer: Softmax Attention with a Forget Gate

• URL: http://arxiv.org/abs/2503.02130v2
• Date: Mon, 31 Mar 2025 19:41:52 GMT
• Title: Forgetting Transformer: Softmax Attention with a Forget Gate
• Authors: Zhixuan Lin, Evgenii Nikishin, Xu Owen He, Aaron Courville,
• Abstract summary: We name this attention mechanism Forgetting Attention and the resulting model the Forgetting Transformer (FoX)<n>FoX outperforms the Transformer on long-context language modeling, length extrapolation, and short-context downstream tasks.<n>FoX is compatible with the FlashAttention algorithm and does not require any positional embeddings.
• Score: 4.484298224007183
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: An essential component of modern recurrent sequence models is the forget gate. While Transformers do not have an explicit recurrent form, we show that a forget gate can be naturally incorporated into Transformers by down-weighting the unnormalized attention scores in a data-dependent way. We name this attention mechanism Forgetting Attention and the resulting model the Forgetting Transformer (FoX). We show that FoX outperforms the Transformer on long-context language modeling, length extrapolation, and short-context downstream tasks, while performing on par with the Transformer on long-context downstream tasks. Moreover, it is compatible with the FlashAttention algorithm and does not require any positional embeddings. Several analyses, including the needle-in-the-haystack test, show that FoX also retains the Transformer's superior long-context capabilities over recurrent sequence models such as Mamba-2, HGRN2, and DeltaNet. We also introduce a "Pro" block design that incorporates some common architectural components in recurrent sequence models and find it significantly improves the performance of both FoX and the Transformer. Our code is available at https://github.com/zhixuan-lin/forgetting-transformer.

Related papers

• Differential Transformer [99.5117269150629]
  Transformer tends to overallocate attention to irrelevant context. We introduce Diff Transformer, which amplifies attention to relevant context while canceling noise. It offers notable advantages in practical applications, such as long-context modeling, key information retrieval, hallucination mitigation, in-context learning, and reduction of activation outliers.
  arXiv  Detail & Related papers  (2024-10-07T17:57:38Z)
• TransformerFAM: Feedback attention is working memory [18.005034679674274]
  We propose a novel Transformer architecture that leverages a feedback loop to enable the network to attend to its own latent representations. TransformerFAM requires no additional weights, enabling seamless integration with pre-trained models.
  arXiv  Detail & Related papers  (2024-04-14T07:43:45Z)
• iTransformer: Inverted Transformers Are Effective for Time Series Forecasting [62.40166958002558]
  We propose iTransformer, which simply applies the attention and feed-forward network on the inverted dimensions. The iTransformer model achieves state-of-the-art on challenging real-world datasets.
  arXiv  Detail & Related papers  (2023-10-10T13:44:09Z)
• Functional Interpolation for Relative Positions Improves Long Context Transformers [86.12843093589]
  We propose a novel functional relative position encoding with progressive, FIRE, to improve Transformer generalization to longer contexts. We theoretically prove that this can represent some of the popular relative position encodings, such as T5's RPE, Alibi, and Kerple. We show that FIRE models have better generalization to longer contexts on both zero-shot language modeling and long text benchmarks.
  arXiv  Detail & Related papers  (2023-10-06T17:59:11Z)
• Fourier Transformer: Fast Long Range Modeling by Removing Sequence Redundancy with FFT Operator [24.690247474891958]
  Fourier Transformer is able to significantly reduce computational costs while retain the ability to inherit from various large pretrained models. Our model achieves state-of-the-art performances among all transformer-based models on the long-range modeling benchmark LRA. For generative seq-to-seq tasks including CNN/DailyMail and ELI5, by inheriting the BART weights our model outperforms the standard BART.
  arXiv  Detail & Related papers  (2023-05-24T12:33:06Z)
• A Length-Extrapolatable Transformer [98.54835576985664]
  We focus on length extrapolation, i.e., training on short texts while evaluating longer sequences. We introduce a relative position embedding to explicitly maximize attention resolution. We evaluate different Transformer variants with language modeling.
  arXiv  Detail & Related papers  (2022-12-20T18:56:20Z)
• Stable, Fast and Accurate: Kernelized Attention with Relative Positional Encoding [63.539333383965726]
  We propose a novel way to accelerate attention calculation for Transformers with relative positional encoding (RPE) Based upon the observation that relative positional encoding forms a Toeplitz matrix, we mathematically show that kernelized attention with RPE can be calculated efficiently using Fast Fourier Transform (FFT)
  arXiv  Detail & Related papers  (2021-06-23T17:51:26Z)
• Glance-and-Gaze Vision Transformer [13.77016463781053]
  We propose a new vision Transformer, named Glance-and-Gaze Transformer (GG-Transformer) It is motivated by the Glance and Gaze behavior of human beings when recognizing objects in natural scenes. We empirically demonstrate our method achieves consistently superior performance over previous state-of-the-art Transformers.
  arXiv  Detail & Related papers  (2021-06-04T06:13:47Z)
• Applying the Transformer to Character-level Transduction [68.91664610425114]
  The transformer has been shown to outperform recurrent neural network-based sequence-to-sequence models in various word-level NLP tasks. We show that with a large enough batch size, the transformer does indeed outperform recurrent models for character-level tasks.
  arXiv  Detail & Related papers  (2020-05-20T17:25:43Z)

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.