- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper studies autoregressive image generation from the perspective of Evidence-KL optimization and Fisher information geometry. It decomposes the CFG signal into orthogonal and parallel components with different roles. The paper finds that the optimal guidance composition varies with guidance strength. It further proposes successive local updates to improve generation quality. The mechanism is validated on multiple autoregressive models.
Critical strengths
-
The analysis of the problem is thorough. The paper is well written and easy to follow. The topic is interesting.
-
The theoretical analysis is solid. The mechanism design is self-consistent.
Critical weaknesses
-
The paper states that weaker guidance emphasizes the orthogonal direction more. It also shows that successive local updates are useful for stronger guidance and for updates that emphasize the orthogonal direction. This seems contradictory.
-
The proposed CFG decomposition is only applied to autoregressive generation. Why? Can it be applied to other settings, such as CFG in diffusion models?
Questions for authors
-
In Figure 4(a) and Figure 5, the smallest FID is achieved by standard CFG with λ=1. Given this, what is the benefit of the proposed decomposition? Where does the improvement come from?
-
How exactly does successive local updates perform each update? Do the repeated updates increase the computational cost? Moreover, the effectiveness of successive local updates is questionable based on the results. Does standard CFG show similar behavior at larger guidance scales?
-
The overall experimental gains are small. The advantage over single guidance is not evident. The practical value seems limited.
-
The visualization does not show clear advantages of the method. In Figure 7(b), the images generated by standard CFG do not collapse. However, λ=15 is already so large that this setting has little practical value.