Yufa Zhou

Yingyu Liang, Zhizhou Sha, Zhenmei Shi et al.

The computational complexity of the self-attention mechanism in popular transformer architectures poses significant challenges for training and inference, and becomes the bottleneck for long inputs. Is it possible to significantly reduce the quadratic time complexity of computing the gradients in multi-layer transformer models? This paper proves that a novel fast approximation method can calculate the gradients in almost linear time $n^{1+o(1)}$ where $n$ is the input sequence length, while it maintains a polynomially small approximation error $1 / \mathrm{poly}(n)$ across the entire model. Our theory holds for general loss functions and when the multi-layer transformer model contains many practical sub-modules, such as residual connection, casual mask, and multi-head attention. By improving the efficiency of gradient computation, we hope that this work will facilitate more effective training and deployment of long-context language models based on our theoretical results.

Xuan Shen, Zhao Song, Yufa Zhou et al.

Diffusion Transformers have emerged as the preeminent models for a wide array of generative tasks, demonstrating superior performance and efficacy across various applications. The promising results come at the cost of slow inference, as each denoising step requires running the whole transformer model with a large amount of parameters. In this paper, we show that performing the full computation of the model at each diffusion step is unnecessary, as some computations can be skipped by lazily reusing the results of previous steps. Furthermore, we show that the lower bound of similarity between outputs at consecutive steps is notably high, and this similarity can be linearly approximated using the inputs. To verify our demonstrations, we propose the \textbf{LazyDiT}, a lazy learning framework that efficiently leverages cached results from earlier steps to skip redundant computations. Specifically, we incorporate lazy learning layers into the model, effectively trained to maximize laziness, enabling dynamic skipping of redundant computations. Experimental results show that LazyDiT outperforms the DDIM sampler across multiple diffusion transformer models at various resolutions. Furthermore, we implement our method on mobile devices, achieving better performance than DDIM with similar latency. Code: https://github.com/shawnricecake/lazydit

Xuan Shen, Chenxia Han, Yufa Zhou et al.

Diffusion transformer-based video generation models (DiTs) have recently attracted widespread attention for their excellent generation quality. However, their computational cost remains a major bottleneck-attention alone accounts for over 80% of total latency, and generating just 8 seconds of 720p video takes tens of minutes-posing serious challenges to practical application and scalability. To address this, we propose the DraftAttention, a training-free framework for the acceleration of video diffusion transformers with dynamic sparse attention on GPUs. We apply down-sampling to each feature map across frames in the compressed latent space, enabling a higher-level receptive field over the latent composed of hundreds of thousands of tokens. The low-resolution draft attention map, derived from draft query and key, exposes redundancy both spatially within each feature map and temporally across frames. We reorder the query, key, and value based on the draft attention map to guide the sparse attention computation in full resolution, and subsequently restore their original order after the attention computation. This reordering enables structured sparsity that aligns with hardware-optimized execution. Our theoretical analysis demonstrates that the low-resolution draft attention closely approximates the full attention, providing reliable guidance for constructing accurate sparse attention. Experimental results show that our method outperforms existing sparse attention approaches in video generation quality and achieves up to 1.75x end-to-end speedup on GPUs. Code: https://github.com/shawnricecake/draft-attention

Yufa Zhou, Shaobo Wang, Xingyu Dong et al.

Directly training Large Language Models (LLMs) for Multi-Agent Systems (MAS) remains challenging due to intricate reward modeling, dynamic agent interactions, and demanding generalization requirements. This paper explores whether post-training techniques, specifically Supervised Fine-Tuning (SFT) and Reinforcement Learning with Verifiable Rewards (RLVR), can effectively $\textit{generalize}$ to multi-agent scenarios. We use economic reasoning as a testbed, leveraging its strong foundations in mathematics and game theory, its demand for structured analytical reasoning, and its relevance to real-world applications such as market design, resource allocation, and policy analysis. We introduce $\textbf{Recon}$ ($\textbf{R}$easoning like an $\textbf{ECON}$omist), a 7B-parameter open-source LLM post-trained on a hand-curated dataset of 2,100 high-quality economic reasoning problems. Comprehensive evaluation on economic reasoning benchmarks and multi-agent games reveals clear improvements in structured reasoning and economic rationality. These results underscore the promise of domain-aligned post-training for enhancing reasoning and agent alignment, shedding light on the roles of SFT and RL in shaping model behavior. Code is available at https://github.com/MasterZhou1/Recon .

Yingyu Liang, Jiangxuan Long, Zhenmei Shi et al.

Large Language Models (LLMs) have shown immense potential in enhancing various aspects of our daily lives, from conversational AI to search and AI assistants. However, their growing capabilities come at the cost of extremely large model sizes, making deployment on edge devices challenging due to memory and computational constraints. This paper introduces a novel approach to LLM weight pruning that directly optimizes for approximating the attention matrix, a core component of transformer architectures. Unlike existing methods that focus on linear approximations, our approach accounts for the non-linear nature of the Softmax attention mechanism. We provide theoretical guarantees for the convergence of our Gradient Descent-based optimization method to a near-optimal pruning mask solution. Our empirical results demonstrate the effectiveness of our non-linear pruning approach in maintaining model performance while significantly reducing computational costs, which is beyond the current state-of-the-art methods, i.e., SparseGPT and Wanda, by a large margin. This work establishes a new theoretical foundation for pruning algorithm design in LLMs, potentially paving the way for more efficient LLM inference on resource-constrained devices.

Yingyu Liang, Zhizhou Sha, Zhenmei Shi et al.

Previous work has demonstrated that attention mechanisms are Turing complete. More recently, it has been shown that a looped 9-layer Transformer can function as a universal programmable computer. In contrast, the multi-layer perceptrons with $\mathsf{ReLU}$ activation ($\mathsf{ReLU}$-$\mathsf{MLP}$), one of the most fundamental components of neural networks, is known to be expressive; specifically, a two-layer neural network is a universal approximator given an exponentially large number of hidden neurons. However, it remains unclear whether a $\mathsf{ReLU}$-$\mathsf{MLP}$ can be made into a universal programmable computer using a practical number of weights. In this work, we provide an affirmative answer that a looped 23-layer $\mathsf{ReLU}$-$\mathsf{MLP}$ is capable of performing the basic necessary operations, more efficiently and effectively functioning as a programmable computer than a looped Transformer. This indicates simple modules have stronger expressive power than previously expected and have not been fully explored. Our work provides insights into the mechanisms of neural networks and demonstrates that complex tasks, such as functioning as a programmable computer, do not necessarily require advanced architectures like Transformers.

Xuan Shen, Zhao Song, Yufa Zhou et al.

Transformers have emerged as the leading architecture in deep learning, proving to be versatile and highly effective across diverse domains beyond language and image processing. However, their impressive performance often incurs high computational costs due to their substantial model size. This paper focuses on compressing decoder-only transformer-based autoregressive models through structural weight pruning to improve the model efficiency while preserving performance for both language and image generation tasks. Specifically, we propose a training-free pruning method that calculates a numerical score with Newton's method for the Attention and MLP modules, respectively. Besides, we further propose another compensation algorithm to recover the pruned model for better performance. To verify the effectiveness of our method, we provide both theoretical support and extensive experiments. Our experiments show that our method achieves state-of-the-art performance with reduced memory usage and faster generation speeds on GPUs.