Comment by Workaccount2 12 hours ago
An update to Gemini diffusion is one of my most eagerly anticipated AI releases. It released to mild fanfare (mostly because you needed to request access to use it), and there has been silence ever since.
Hopefully it's not more Google abandonware, because it was wicked fast and a delight to use
Diffusion LMs do seem to be able to get more out of the same data. In a world where we are already training transformer based LLMs on all text available, diffusion LMs ability to continue learning on a fixed set of data may be able to outperform transformers
Perhaps it’s an issue is that text often has directionality.
Latency may be better, but throughput (the thing companies care about) may be the same or worse, since every step the entire diffusion window has to be passed through the model. With AR models only the most recent token goes through, which is much more compute efficient allowing you to be memory bound. Trade off with these models is more than one token per forward pass, but idk the point where that becomes worth it (probably depends on model and diffusion window size)
Yes but you can also do the same thing with autoregressive models just by making them smaller. This tradeoff always exists, the question is whether the Pareto curve for diffusion models ever crosses or dominates the best autoregressive option at the same throughput (or quality).
It's not a very promising direction because autoregressive LLMs still deliver better output quality per model weight, as a rule.
Now, is it possible that a model can combine advantages of both? Combine fast generation and multidirectional causality of diffusion with precision, capabilities and generalization of autoregression?
Maybe. This paper is research in that direction. So far, it's not a clear upgrade over autoregressive LLMs.