DeepSeek speculative decoding framework DSpark went live June 27 on V4-Flash and V4-Pro, reporting up to 85 percent faster ...

Deploying DFlash block diffusion on NVIDIA hardware accelerates autoregressive LLMs during latency-sensitive inference.

Google has released DiffusionGemma, an experimental language model that generates text using a diffusion-based method, producing blocks of 256 tokens at once rather than generating text word by word.

Abstract: Automatic speech recognition (ASR) systems often rely on autoregressive (AR) Transformer decoder architectures, which limit efficient inference parallelization due to their sequential nature ...

We introduce OneCAT, a unified multimodal model that seamlessly integrates understanding, generation, and editing within a novel, pure decoder-only transformer architecture. Our framework uniquely ...

Credit: VentureBeat made with OpenAI ChatGPT-Images-2.0 In a significant shift toward local-first privacy infrastructure, OpenAI has released Privacy Filter, a ...

The AI image generation market has had an uncontested leader for months. Google's Nano Banana family of models has set the standard for quality, speed, and commercial adoption, while competitors from ...

Midjourney is rolling out an early version of its new V8 model for community testing. Image generation is reportedly much faster and more detailed, but some features cost four times as much.

For visual generation, discrete autoregressive models often struggle with poor tokenizer reconstruction, difficulties in sampling from large vocabularies, and slow token-by-token generation speeds. We ...