- Rating
- 3
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
The paper proposes a training-free sampler for uniform-state diffusion LLMs, evaluated on DiffusionGemma. The motivation is that a position can stop changing across denoising steps while its distribution is still broad, and current samplers keep waiting for it to sharpen. The authors define a Convergence Probability , motivated by the method of types, turn it into a convergence entropy , and, for argmax-stable positions, use in place of the token entropy in the EB-Sampler and in adaptive stopping. On eight benchmarks the method reduces denoising steps relative to the native EB-Sampler, the largest reduction among Fast-dLLM, KLASS and DAPD, while mildly harming quality.
Critical strengths
1.Convergence vs. confidence.
A token can stop changing between steps while its distribution is still not sharp, and confidence or entropy rules miss this. The re-noising argument is a nice addition: uniform-state models can undo a wrong accept, so a looser gate makes sense here, unlike in masked models.
- Consistent step reduction without training.
The method needs no retraining or model changes, and reduces the number of denoising steps on all eight benchmarks.
Critical weaknesses
- The comparison protocol and the accuracy-steps trade-off need to be clearer.
Section 2.2 describes DiffusionGemma's sampler as EB-Sampler plus temperature annealing and adaptive stopping, and Section 4.1 compares against the "native EB-Sampler of DiffusionGemma". It is not stated which of these components (annealing, adaptive stopping, re-noising) each baseline keeps. Fast-dLLM, KLASS and DAPD were designed for masked models with irreversible commits, so using re-noising for some methods but not others would confound the comparison. Also, replaces the entropy in the stopping rule as well, so part of the step reduction may come from stopping earlier rather than accepting earlier.
On accuracy, the method is below the EB-Sampler on 7 of 8 benchmarks. In the proposed sampler is fixed at 0.1 (and is tied to it), while the EB-Sampler can trade steps for accuracy by sweeping . With a single point per method, it is hard to tell whether the new sampler improves the trade-off or moves along it.
A Pareto plot of accuracy vs. steps, sweeping for both the EB-Sampler and the proposed sampler under the same protocol, together with a clear statement of which components each baseline uses, would address most of this. If the proposed curve lies above the EB-Sampler curve, the contribution would be much clearer and I would raise my score.
- The method-of-types link and the "0 new hyperparameters" claim are not supported.
Lemma 1 holds for types of length- sequences. At a type is one-hot, while is soft, so the lemma does not directly make a probability of convergence. In practice is the KL between steps mapped to .
Hence is a KL threshold () rather than a majority rule, and follows from the same choice while capping the cost of every argmax-stable position (). With the two-step stability window and the KL direction, the method has design parameters, so "0 new" in Table 1 is not accurate; Section 3.2 itself notes the threshold could be tuned.
I suggest revisiting how is presented and how it relates to KL, and clarifying the method-of-types connection: where Lemma 1 applies, what changes at , and in what sense remains a likelihood of convergence. The threshold and should be justified or swept, and an ablation against argmax stability alone and a plain KL threshold would show which part drives the step reduction.
Questions for authors
Questions whose answers could change my assessment:
-
Accuracy-vs-steps (Pareto) curves for the EB-Sampler and your sampler, sweeping under the same protocol?
-
Which of annealing, adaptive stopping and re-noising does each baseline use? How much of the saving comes from earlier stopping vs. earlier acceptance?
-
A sweep of the threshold and , and an ablation with argmax stability alone?
-
How often does the converged-but-broad case (Sec. 3.1) occur across the benchmarks?
Clarifications:
-
Is KL computed on tempered or raw logits? Annealing changes even when the logits do not.
-
Can the anchor effect (Sec. 3.3) be measured beyond heatmaps?
-
How does the method relate to the LLaDA 2.1 sampler and to remasking samplers?
-
In Fig. 3, which positions are accepted, and how is stability decided?
-
Results on another uniform-state or masked model, to support Sec. 3.4?
Minor suggestions:
• Fig. 1: panel (a) is unrelated to the method - clear why presenting it; the 0-100 axis in (b) hides the differences.
• Table 2: mark the best result per row and define TPF in the caption.
Discussion Reminder
We sincerely thank you for taking the time to review our submission. We have provided detailed responses to your concerns in the comments above and hope they address your questions.
As the discussion period is coming to an end, we would greatly appreciate any further feedback or concerns you may have.
Thank you again for your time and consideration.