Related papers: SparseD: Sparse Attention for Diffusion Language M…

In recent years, Large Language Models (LLMs) have demonstrated remarkable capabilities across a wide array of text-centric tasks. However, their `large' scale introduces significant computational and storage challenges, particularly in…

This study introduces bifurcated attention, a method designed to enhance language model inference in shared-context batch decoding scenarios. Our approach addresses the challenge of redundant memory IO costs, a critical factor contributing…

Diffusion Language Models (DLMs) are often advertised as enabling parallel token generation, yet practical fast DLMs frequently converge to left-to-right, autoregressive (AR)-like decoding dynamics. In contrast, genuinely non-AR generation…

Speculative decoding (SD) accelerates Large Language Model (LLM) generation by using an efficient draft model to propose the next few tokens, which are verified by the LLM in a single forward call, reducing latency while preserving its…

Diffusion large language models (dLLMs) are emerging as promising alternatives to autoregressive (AR) LLMs. Recently, this paradigm has been extended to multimodal tasks, leading to the development of diffusion multimodal large language…

Most multi-agent systems rely exclusively on autoregressive language models (ARMs) that are based on sequential generation. Although effective for fluent text, ARMs limit global reasoning and plan revision. On the other hand, Discrete…

Autoregressive (AR) models remain the standard for natural language generation but still suffer from high latency due to strictly sequential decoding. Recent diffusion-inspired approaches, such as LlaDA and Dream, mitigate this by…

Self-attention has been a huge success for many downstream tasks in NLP, which led to exploration of applying self-attention to speech problems as well. The efficacy of self-attention in speech applications, however, seems not fully blown…

Current hierarchical attention methods, such as NSA and InfLLMv2, select the top-k relevant key-value (KV) blocks based on coarse attention scores and subsequently apply fine-grained softmax attention on the selected tokens. However, the…

A new line of research for feature selection based on neural networks has recently emerged. Despite its superiority to classical methods, it requires many training iterations to converge and detect informative features. The computational…

Transformer-based models are popularly used in natural language processing (NLP). Its core component, self-attention, has aroused widespread interest. To understand the self-attention mechanism, a direct method is to visualize the attention…

The scalability of high-fidelity video diffusion models (VDMs) is constrained by two key sources of redundancy: the quadratic complexity of global spatio-temporal attention and the computational overhead of long iterative denoising…

Emerging Large Language Model (LLM) applications require long input context in order to perform complex tasks like document analysis and code generation. For these long context length applications, the length of the input prompt poses a…

Masked Diffusion Models (MDMs) offer a promising alternative to autoregressive language models by enabling parallel token generation and bidirectional context modeling. However, their inference speed is significantly limited by the…

We propose \textbf{MoE-DiffuSeq}, a diffusion-based framework for efficient long-form text generation that integrates sparse attention with a Mixture-of-Experts (MoE) architecture. Existing sequence diffusion models suffer from prohibitive…

Diffusion Transformers (DiTs) dominate video generation but their high computational cost severely limits real-world applicability, usually requiring tens of minutes to generate a few seconds of video even on high-performance GPUs. This…

Sparse attention improves LLM inference efficiency by selecting a subset of key-value entries, but at the cost of potential accuracy degradation. In particular, omitting critical KV entries can induce substantial errors in model outputs.…

Diffusion models (DMs) have shown promising results on single-image super-resolution and other image-to-image translation tasks. Benefiting from more computational resources and longer inference times, they are able to yield more realistic…

The quadratic computational complexity of self-attention remains a fundamental bottleneck for scaling Large Language Models (LLMs) to long contexts, particularly during the pre-filling phase. In this paper, we rethink the causal attention…

Diffusion models demonstrate outstanding performance in image generation, but their multi-step inference mechanism requires immense computational cost. Previous works accelerate inference by leveraging layer or token cache techniques to…