News for hackers. Stuff that matters.
Re:Show HN: I invented a new generative model and got accept... by diyer22 on Saturday October 11, 2025 @02:24AM (#45547068)
Even in the earliest stages of the DDN concept, we had already decided to pass features down to the next layer.

I never even ran an ablation that disabled the stem features; I assume the network would still train without them, but since the previous layer has already computed the features, it would be wasteful not to reuse them. Retaining the stem features also lets DDN adopt the more efficient single-shot-generator architecture.

Another deeper reason is that, unlike diffusion models, DDN does not need the Markov-chain property between adjacent layers.

Re:Show HN: I invented a new generative model and got accept... by cttet on Saturday October 11, 2025 @02:28AM (#45547078)
Thanks! Really like your intuition!