undefined | Better HN

0 pointsAlexCoventry10mo ago0 comments

> it can adjust a whole block of tokens when it encounters some kind of disjunction.

This is true in principle for general diffusion models, but I don't think it's true for the noise model they use in Mercury (at least, going by a couple of academic papers authored by the Inception co-founders.) Their model generates noise by masking a token, and once it's masked, it stays masked. So the reverse-diffusion gets to decide on the contents of a masked token once, and after that it's fixed.

0 comments

freeqaz10mo ago

Here are two papers linked from Inception's site:

1. Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution - https://arxiv.org/abs/2310.16834

2. Simple and Effective Masked Diffusion Language Models - https://arxiv.org/abs/2406.07524

AlexCoventryOP10mo ago

Thanks, yes, I was thinking specifically of "Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution". They actually consider two noise distributions: one with uniform sampling for each noised token position, and one with a terminal masking (the Q^{uniform} and Q^{absorb}.) However, the terminal-masking system is clearly superior in their benchmarks.

https://arxiv.org/pdf/2310.16834#page=6

macleginn10mo ago

The exact types of path dependencies in inference on text-diffusion models look like an interesting research project.

AlexCoventryOP10mo ago

Yes, the problem is coming up with a noise model where reverse diffusion is tractable.

XenophileJKO10mo ago

Thank you, I'll have to read the papers. I don't think I have read theirs.

j / k navigate · click thread line to collapse

0 pointsAlexCoventry10mo ago0 comments

> it can adjust a whole block of tokens when it encounters some kind of disjunction.

0 comments

freeqaz10mo ago

Here are two papers linked from Inception's site:

1. Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution - https://arxiv.org/abs/2310.16834

2. Simple and Effective Masked Diffusion Language Models - https://arxiv.org/abs/2406.07524

AlexCoventryOP10mo ago

https://arxiv.org/pdf/2310.16834#page=6

macleginn10mo ago

The exact types of path dependencies in inference on text-diffusion models look like an interesting research project.

AlexCoventryOP10mo ago

Yes, the problem is coming up with a noise model where reverse diffusion is tractable.

XenophileJKO10mo ago

Thank you, I'll have to read the papers. I don't think I have read theirs.

j / k navigate · click thread line to collapse