8 citations · 11 across the 26 of their papers we have counts for
14 papers · 1 filter
LeapQuant: Efficient Linear Attention with Accurate Recurrent State Quantization
Yi Pan, Haocheng Xi, Kan Zhu +10
Recent LLMs increasingly adopt hybrid designs that replace standard attention with linear attention, such as Gated DeltaNet (GDN) and Kimi Delta Attention (KDA). Although they comp…
Video DeltaNet: A Video-Native Hybrid Attention for Livestream Video Generation
Haocheng Xi, Yiming Xie, Hexu Zhao +8
Video diffusion models repeatedly process long spatiotemporal token sequences during denoising, making attention a major computational bottleneck. Linear attention offers an appeal…
FourTune: Towards Fully 4-Bit Efficient Post-Training for Diffusion Models
Bowen Xue, Zihan Min, Xingyang Li +8
Diffusion models have become a dominant paradigm for high-quality generative modeling, while post-training is essential for adapting them to diverse downstream applications. Howeve…
Quant VideoGen: Auto-Regressive Long Video Generation via 2-Bit KV-Cache Quantization
Haocheng Xi, Shuo Yang, Yilong Zhao +13
Despite rapid progress in autoregressive video diffusion, an emerging system algorithm bottleneck limits both deployability and generation capability: KV cache memory. In autoregre…
Jet-RL: Enabling On-Policy FP8 Reinforcement Learning with Unified Training and Rollout Precision Flow
Haocheng Xi, Charlie Ruan, Peiyuan Liao +7
Reinforcement learning (RL) is essential for enhancing the complex reasoning capabilities of large language models (LLMs). However, existing RL training pipelines are computational…
SciML Agents: Write the Solver, Not the Solution
Saarth Gaonkar, Xiang Zheng, Haocheng Xi +5
Recent work in scientific machine learning aims to tackle scientific tasks directly by predicting target values with neural networks (e.g., physics-informed neural networks, neural…
SLA: Beyond Sparsity in Diffusion Transformers via Fine-Tunable Sparse-Linear Attention
Jintao Zhang, Haoxu Wang, Kai Jiang +10
In Diffusion Transformer (DiT) models, particularly for video generation, attention latency is a major bottleneck due to the long sequence length and the quadratic complexity. We f…
XQuant: Breaking the Memory Wall for LLM Inference with KV Cache Rematerialization
Aditya Tomar, Coleman Hooper, Minjae Lee +7
Although LLM inference has emerged as a critical workload for many downstream applications, efficiently inferring LLMs is challenging due to the substantial memory footprint and ba…
QuantSpec: Self-Speculative Decoding with Hierarchical Quantized KV Cache
Rishabh Tiwari, Haocheng Xi, Aditya Tomar +7
Large Language Models (LLMs) are increasingly being deployed on edge devices for long-context settings, creating a growing need for fast and efficient long-context inference. In th…
Oscillation-Reduced MXFP4 Training for Vision Transformers
Yuxiang Chen, Haocheng Xi, Jun Zhu +1
Pre-training Transformers in FP4 precision is becoming a promising approach to gain substantial speedup, but it comes with a considerable loss of accuracy. Microscaling (MX) data f…
SpargeAttention: Accurate and Training-free Sparse Attention Accelerating Any Model Inference
Jintao Zhang, Chendong Xiang, Haofeng Huang +4
An efficient attention implementation is essential for large models due to its quadratic time complexity. Fortunately, attention commonly exhibits sparsity, i.e., many values in th…
COAT: Compressing Optimizer states and Activation for Memory-Efficient FP8 Training
Haocheng Xi, Han Cai, Ligeng Zhu +4
FP8 training has emerged as a promising method for improving training efficiency. Existing frameworks accelerate training by applying FP8 computation to linear layers while leaving…
Jetfire: Efficient and Accurate Transformer Pretraining with INT8 Data Flow and Per-Block Quantization
Haocheng Xi, Yuxiang Chen, Kang Zhao +3
Pretraining transformers are generally time-consuming. Fully quantized training (FQT) is a promising approach to speed up pretraining. However, most FQT methods adopt a quantize-co…
Training Transformers with 4-bit Integers
Haocheng Xi, Changhao Li, Jianfei Chen +1
Quantizing the activation, weight, and gradient to 4-bit is promising to accelerate neural network training. However, existing 4-bit training methods require custom numerical forma…