Google DeepMind has developed DiffusionGemma, a text diffusion model that was created by retrofitting Gemma 4. This approach required less than 10% of the original training budget and allows for parallel generation of 256 tokens, achieving speeds of approximately 1,500 tokens per second. While faster than traditional autoregressive models, DiffusionGemma's performance on reasoning tasks still lags behind its predecessor. AI
IMPACT This approach could significantly reduce the computational cost and time required to develop new text generation models.
RANK_REASON The item describes a new research approach to creating a text diffusion model by retrofitting an existing model, rather than training from scratch. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
