Trends · 研究趋势
数据来自 Hugging Face 论文的 AI 提取关键词,按月统计研究方向的增长与热度。
Jianuo Huang, Yaojie Zhang, Qituan Zhang +3 authors
Domino is a speculative decoding framework that improves LLM inference speed by decoupling causal dependency modeling from autoregressive drafting through a parallel backbone and lightweight causal refinement head, achieving significant speedups in both end-to-end execution and throughput.
Mohamad Zbib, Mohamad Bazzi, Ammar Mohanna +2 authors
Speculative decoding effectiveness depends on draft model training data alignment with downstream tasks, with specialized drafters performing better when combined through confidence-based routing rather than simple averaging.
Subham Sekhar Sahoo, Lingjie Chen, Khiem Pham +14 authors
Diffusion-augmented autoregressive language models use parallel token sampling via distilled diffusion weights and a specialized sampler to accelerate inference without quality loss or draft models.
Jingyu Liu, Xin Dong, Zhifan Ye +6 authors
TiDAR combines diffusion and autoregressive models to achieve high throughput and quality in language generation.
Hong Liu, Jiaqi Zhang, Chao Wang +13 authors
Embedding scaling offers superior sparsity scaling compared to expert scaling in large language models, enabling efficient inference through system optimizations and speculative decoding.
Jian Chen, Yesheng Liang, Zhijian Liu
DFlash is a speculative decoding framework that uses a lightweight block diffusion model for parallel token drafting, achieving significant speedup over existing autoregressive methods while maintaining high-quality outputs.
Team Seedream, Yunpeng Chen, Yu Gao +47 authors
Seedream 4.0 is a high-performance multimodal image generation system that integrates text-to-image synthesis, image editing, and multi-image composition using a diffusion transformer and VAE, achieving state-of-the-art results with efficient training and inference.
Hao Zhang, Yiming Hu, Yong Wang +3 authors
Speculative decoding with adaptive block size selection improves inference efficiency by predicting optimal block sizes from prefilling representations, achieving significant speedup with minimal overhead.
Tianle Cai, Yuhong Li, Zhengyang Geng +4 authors
Medusa enhances Large Language Model inference by adding parallel decoding heads and using tree-based attention to predict multiple tokens simultaneously, achieving significant speedup with minimal latency.
Sanchit Gandhi, Patrick von Platen, Alexander M. Rush
Distil-Whisper, a smaller and faster variant of the Whisper model, achieves nearly the same performance with fewer resources and is optimized for low-latency environments.
Junxiong Wang, Daniele Paliotta, Avner May +2 authors
Large-scale Transformer models can be distilled into linear RNNs using attention layer weights, achieving competitive performance with enhanced deployment efficiency.
Nikhil Bhendawade, Irina Belousova, Qichen Fu +3 authors
Speculative Streaming is a method that incorporates drafting into a target language model, enhancing decoding speed without quality loss while using fewer parameters.
DiffusionGemma Team, Adrien Ali Taïga, James Assiene +41 authors
DiffusionGemma is a fine-tuned mixture-of-experts language model that uses discrete diffusion to generate text blocks in parallel, achieving high speed while preserving capabilities like multimodal inputs and reasoning.
Xin Cheng, Xingkai Yu, Chenze Shao +30 authors
DSpark enhances LLM inference speed by combining parallel draft generation with adaptive verification that reduces waste and improves throughput in high-concurrency settings.
Benjamin Spector, Chris Re
A staged speculative decoding algorithm enhances LLM inference efficiency in small-batch, on-device settings by restructuring speculative batches and adding a second decoding stage.
Geonmo Gu, Byeongho Heo, HeeJae Jun +4 authors
Verification-Aware Training improves draft models for speculative decoding by simulating sequential verification during training and adapting loss weights to acceptance patterns.
北京市昌平区探索星信息技术及软件开发工作室
京ICP备2026059466号