LLaDA2.0-mini
PulseAugur coverage of LLaDA2.0-mini — every cluster mentioning LLaDA2.0-mini across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
MoE Diffusion Decoding Bottleneck Identified and Addressed
Researchers have identified a significant bottleneck in the decoding process of Mixture-of-Experts (MoE) diffusion language models. They found that after weight quantization, the per-token sampling and stopping pipeline…
-
LLaDA-Image sets new open-source SOTA for image generation
Researchers have introduced LLaDA-Image, a novel framework for generating high-quality images using a 6B Diffusion Transformer trained from scratch. This model leverages image-only pre-training and a specialized optimiz…
-
Masked diffusion LLMs use EoS tokens for hidden reasoning
Researchers have discovered that masked diffusion large language models (LLMs) can leverage end-of-sequence (EoS) tokens for hidden reasoning, enhancing their performance on complex tasks. By padding answers with EoS to…
-
LLM inference and reasoning techniques advance with new research and hardware
Researchers are exploring novel methods to enhance the efficiency and reasoning capabilities of large language models (LLMs). Google Research is developing techniques to train LLMs to reason in a Bayesian manner, improv…