Understand this story
On November 5, 2026, Deng Zhijie, an associate professor at Shanghai Jiao Tong University, gave a presentation at CNCC 2026, systematically reviewing acceleration methods for diffusion language models, from the foundational LLaDA and Dream approaches to cross-block parallelism, post-training, and inference system optimization. He introduced that Fast-dLLM requires no additional training and, under the settings of LLaDA-Instruct, GSM8K 5-shot, a single A100, and a batch size of 1, increases speed from 6.7 to 54.4 TPS, with accuracy changing from 79.3% to 78.5%. Deng Zhijie also introduced explorations of related methods in streaming video, speech restoration, multimodal, and agent applications. Currently, this presentation has been reported by Leiphone, and there are no further developments yet.
AI-generated from public reports
Developments
1 confirmed developments
Deng Zhijie gives a report on diffusion large language model methods and applications at CNCC 2026
Source:English text is being prepared.
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 5
Leiphone
Shanghai Jiao Tong University's Deng Zhijie: How Diffusion Language Models Can Balance Generation Speed and Quality
At CNCC 2026, Deng Zhijie, an associate professor at Shanghai Jiao Tong University, systematically reviewed acceleration methods for diffusion language models, from the foundational LLaDA and Dream approaches to cross-block parallelism, post-training, and inference system optimization. Among them, Fast-dLLM requires no additional training and, under the settings of LLaDA-Instruct, GSM8K 5-shot, a single A100, and batch size 1, increased speed from 6.7 to 54.4 TPS, while accuracy changed from 79.3% to 78.5%. He also introduced explorations of related methods in streaming video, speech restoration, multimodal, and agent applications.
AI Trends