Related papers: Energy-Gated Attention and Wavelet Positional Enco…
Transformer attention architectures, similar to those developed for natural language processing, have recently proved efficient also in vision, either in conjunction with or as a replacement for convolutional layers. Typically, visual…
The computational burden of attention in long-context language models has motivated two largely independent lines of work: sparse attention mechanisms that reduce complexity by attending to selected tokens, and gated attention variants that…
The self-attention module is a key component of Transformer-based models, wherein each token pays attention to every other token. Recent studies have shown that these heads exhibit syntactic, semantic, or local behaviour. Some studies have…
Transformer-based models have demonstrated considerable potential for source code modeling tasks in software engineering. However, they are limited by their dependence solely on automatic self-attention weight learning mechanisms. Previous…
In this work, we investigate the positional encoding methods used in language pre-training (e.g., BERT) and identify several problems in the existing formulations. First, we show that in the absolute positional encoding, the addition…
In the realm of large-scale language models, a significant challenge arises when extrapolating sequences beyond the maximum allowable length. This is because the model's position embedding mechanisms are limited to positions encountered…
We use a mean-field-based transformer model to theoretically investigate how auxiliary variables, such as positional encoding, prevent mode collapse of self-attention mechanisms. The use of mean-field transformers to analyze the properties…
We explain why directly changing the prior can be a surprisingly ineffective mechanism for incorporating inductive biases into VAEs, and introduce a simple and effective alternative approach: Intermediary Latent Space VAEs(InteL-VAEs).…
This paper investigates fake news detection as a downstream evaluation of Transformer representations, benchmarking encoder-only and decoder-only pre-trained models (BERT, GPT-2, Transformer-XL) as frozen embedders paired with lightweight…
Transformer architectures, underpinned by the self-attention mechanism, have achieved state-of-the-art results across numerous natural language processing (NLP) tasks by effectively modeling long-range dependencies. However, the…
Attention-based neural networks, such as Transformers, have become ubiquitous in numerous applications, including computer vision, natural language processing, and time-series analysis. In all kinds of attention networks, the attention maps…
Attention mechanism has become the dominant module in natural language processing models. It is computationally intensive and depends on massive power-hungry multiplications. In this paper, we rethink variants of attention mechanism from…
Standard attention mechanisms in transformers employ static token representations that remain unchanged across all pair-wise computations in each layer. This limits their representational alignment with the potentially diverse relational…
We present the Condensate Theorem: attention sparsity is a learned topological property, not an architectural constraint. Through empirical analysis of trained language models, we find that attention mass concentrates on a distinct…
Transformers are increasingly dominating multi-modal reasoning tasks, such as visual question answering, achieving state-of-the-art results thanks to their ability to contextualize information using the self-attention and co-attention…
Decomposing model activations into interpretable components is a key open problem in mechanistic interpretability. Sparse autoencoders (SAEs) are a popular method for decomposing the internal activations of trained transformers into sparse,…
Vision Transformers has demonstrated competitive performance on computer vision tasks benefiting from their ability to capture long-range dependencies with multi-head self-attention modules and multi-layer perceptron. However, calculating…
Many positional encodings (PEs) are designed to exhibit long-term decay, based on an entrenched and long-standing inductive opinion: tokens farther away from the current position carry less relevant information. We argue that long-term…
Unsupervised pre-training on vast amounts of graph data is critical in real-world applications wherein labeled data is limited, such as molecule properties prediction or materials science. Existing approaches pre-train models for specific…
We introduce exclusive self attention (XSA), a simple modification of self attention (SA) that improves Transformer's sequence modeling performance. The key idea is to constrain attention to capture only information orthogonal to the…