Long-Context Generalization with Sparse Attention
Transformer-based architectures traditionally employ softmax to compute attention weights, which produces dense distributions over all tokens in a sequence. While effective in many settings, this density has been shown to be detrimental for tasks that demand precise focus on fixed-size patterns: as sequence length increases, non-informative tokens accumulate attention probability mass, leading to dispersion and representational collapse. We show in this paper that sparse attention mechanisms using $\alpha$-entmax can avoid these issues, due to their ability to assign exact zeros to irrelevant tokens. Furthermore, we introduce Adaptive-Scalable Entmax (ASEntmax), which endows $\alpha$-entmax with a learnable temperature parameter, allowing the attention distribution to interpolate between sparse (pattern-focused) and dense (softmax-like) regimes. Finally, we show that the ability to locate and generalize fixed-size patterns can be further improved through a careful design of position encodings, which impacts both dense and sparse attention methods. By integrating ASEntmax into standard transformer layers alongside proper positional encodings, we show that our models greatly outperform softmax, scalable softmax, and fixed-temperature $\alpha$-entmax baselines on long-context generalization.
Code (0)
등록된 구현이 없습니다.
Methods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
ReSSFormer: A Recursive Sparse Structured Transformer for Scalable and Long-Context Reasoning
While Transformer architectures have demonstrated impressive scalability across domains, they continue to face challenges in long-context reasoning, computational efficiency, and structural generalization - largely due t…
Computational EfficiencyEvery Token Counts: Generalizing 16M Ultra-Long Context in Large Language Models
This work explores the challenge of building ``Machines that Can Remember'', framing long-term memory as the problem of efficient ultra-long context modeling. We argue that this requires three key properties: \textbf{spa…
MTraining: Distributed Dynamic Sparse Attention for Efficient Ultra-Long Context Training
The adoption of long context windows has become a standard feature in Large Language Models (LLMs), as extended contexts significantly enhance their capacity for complex reasoning and broaden their applicability across d…
$π$-Attention: Online Efficient Sparse Transformers for Long-Context Modeling
Sparse attention is crucial in long-context Transformers, which restricts each token to a limited neighborhood and thereby reduces the quadratic cost of full self-attention. Local windows capture nearby context effective…
Long-range modelingUnderstanding and Improving Length Generalization in Hierarchical Sparse Attention Models
Effectively processing long contexts is a critical challenge for language models. While standard Transformers are limited by quadratic complexity and poor length extrapolation, alternative architectures like sliding wind…