- Rating
- 2
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper introduces , a self-alignment training method for pixel-space diffusion models. It replaces the EMA targets of the prior latent-space works with online targets, while introducing additional variance regularization techniques to mitigate training difficulties. The experiments on ImageNet-256 show that its improvement on model performance is decent.
Critical strengths
- The motivation is easy to follow, and the overall presentation is clear.
- The performance of SiamFlow on small-size model is on par with PixelREPA, which relies on external models.
Critical weaknesses
My primary concern is that, given the current state of the paper, the main motivation appears to be simply attempting self-alignment training in pixel space and introducing several heuristics to improve the FID score. And the resulting performance gains seem marginal and decrease monotonically as the model parameter count increases (Table 3).
As pointed out by the authors, certain techniques from latent-space self-alignment methods do not directly transfer to pixel-space training. Specifically, the introduction mentions that prior latent-space works favor EMA targets over online targets. Given this, the paper shoul focus more on explaining the role of EMA targets and why they can be simplified using variance regularization. Additionally, does this regularization also benefit latent-space methods? What are the key takeaways for self-alignment training in pixel space, and how much further can these methods be simplified? I'm not asking for more experiments but providing potential directions for refining the paper.
Minors on writings and presentations
- To properly position the paper within the literature, I suggest providing a deeper discussion of the technical differences between the proposed method and relevant prior work, rather than merely introducing them sequentially.
- In line 159, I suggest using distinct symbols like and to denote different timesteps rather than using with subscripts. In line 180, the authors should define the meaning of and . For completeness, it would also be helpful to explain the core intuition behind self-alignment and why it works. Finally, Section 6.3 is currently confusingly written and should be revised to eliminate ambiguity.
Questions for authors
- What does "Gap" signify in Table 1?
- Could you clarify what is meant by "cross-noise retrieval"?
- Compared to the JiT baseline, the performance gains decrease noticeably on larger models. Considering this, how much does SiamFlow improve over latent-space self-alignment approaches (such as SRA or Self-Flow) when scaled to larger models (e.g., 676M parameters)?
I thank the authors for their detailed responses. After carefully reading the rebuttal, the additional experiments address several of my methodological questions. However, they do not substantially address my concerns.
Feature alignment, token standardization, etc, are established ideas, and the contribution is mainly careful adaptation to pixel-space generative training based on JiT. The contribution is therefore primarily an empirical adaptation to pixel-space self-alignment. The gains are also modest. With evidence limited to JiT on ImageNet, the broader insights of the paper remains unclear.
The current state of the paper reads more like an expanded rebuttal than a coherent study, making it difficult to identify the primary contribution and takeaway.
Therefore, I decide to maintain my current weak rejection assessment.