- Rating
- 3
- Confidence
- 3
- Informativeness
- 3
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
T-DIG formulates image-set diversity as generation conditioned on a user-written axis of variation. Two LoRA-adapted language policies produce prompt sets or editing instructions for frozen image generators/editors. GRPO optimizes a normalized TPIPS-conditioned Vendi reward with baseline-relative alignment and preference penalties and a length penalty; the editing policy also receives a DreamSim-based drift penalty. Training axes are primarily generated by a VLM using human-written pilot responses as examples, while the test requests come from user interactions. The paper reports diversity improvements across four open-weight generators, transfer to two API generators, and human comparisons, with some alignment and preference tradeoffs.
Critical strengths
The task formulation addresses differences in users' intentions by allowing free-form variation axes for the same prompt without requiring a fixed attribute taxonomy. The text-policy interface also makes the approach applicable to frozen renderers from different model families. Table 3 reports diversity gains across four generators, while Table 1 shows that people generally judge the sets produced at α=0.5 as more varied along the requested axis than those from the external baselines. The 50-prompt API transfer experiment provides additional, although limited, evidence of applicability beyond open-weight models. The comparison with the untrained policy makes the contribution of training more credible. This control receives the same axis-conditioned instruction, yet the α=1 adapter adds approximately 0.063–0.114 normalized TPIPS-Vendi across the four generators in Table 3. Table 2 also shows why optimizing diversity alone is insufficient: it raises the diversity score to 0.617 while reducing VQAScore to 0.228. The alignment, length, and preference penalties address different observed failure modes, providing useful evidence for the reward design rather than only a final aggregate result.
Critical weaknesses
Incomplete reporting limits the assessment of the human evidence. Section 4.4 describes 80 participants and 400 test pairs, whereas RQ5 reports 53 participants and 415 pairs without explaining their relationship. Table 1 also omits cell-level sample sizes, win/loss/tie counts, and a precise definition of the sign test underlying its significance statement. This matters particularly for the joint diversity-and-alignment criterion: a set must be more varied and at least as well aligned to count as a joint win, so neither method may satisfy the criterion. Consequently, the reported joint win rates of 47.6%–58.3% against the untrained policy cannot, by themselves, establish or refute statistical significance. The authors should specify the null hypothesis, denominator, handling of ties and cases where neither method jointly wins, and which columns the caption's significance statement covers. Until these details are provided, the strength of the claimed human-evaluation advantage remains uncertain. The evidence is also less complete for preserving required content outside the requested axis and for iterative editing than for increasing diversity. TPIPS-Vendi is both a training objective and the principal automatic diversity metric; the human comparisons are helpful, but do not directly quantify retention of individual prompt constraints. Tables 1–3 do not clearly separate generation from editing or characterize drift over successive rounds, although Appendix A.2 reports unintended hair-color changes and accumulating editing artifacts. At α=1, Table 3 additionally shows HPSv3 decreases of about 6.1% for Qwen-Image and 10.4% for FLUX.2 [klein] relative to i.i.d. sampling. Thus, the current evidence supports a useful diversity/quality tradeoff more strongly than a general preservation claim.
Questions for authors
- Please explain how Section 4.4's 80 participants/400 pairs relate to RQ5's 53 participants/415 pairs, including cohorts, rounds, and exclusions. For Table 1, please provide cell-level sample sizes, win/loss/tie counts or the underlying rating differences, confidence intervals, and the precise sign-test definition. Which comparisons does the caption's significance statement cover, and how are cases where neither set jointly wins handled?
- Please separate generation and editing sample sizes/results and specify the editing ν and δ and the construction of its baseline-relative thresholds. Under matched axes and budgets, how do trained and untrained editing policies compare on target-axis variation, non-target content preservation, and drift over successive rounds?
- What motivated α=0.5 for human evaluation and α=1 for the main automatic comparison? Can you report the automatic diversity, alignment, and preference results at the human-study setting as well, so that the trained-versus-untrained tradeoff can be compared at the same operating point and noise protocol?
4 comments
Weakness 2: Preserving content outside the requested axis
We agree, and we do not claim that T-DIG leaves all non-target content unchanged: our claim is that it raises diversity along the requested axis while preserving base-prompt alignment (L50–51). However, we do briefly discuss this as a motivation to add the length penalty (L200–204). All images in a set also share the same noise latent, so the variation comes only from the prompt. That said, we agree that off-axis change should be quantified, so we measure it for both generation and editing. We also have more discussions on how the strength control which the user sets, controls for this a bit, and some diversity is good because it helps with creative exploration (but more on this below).
Generation: We compute the TPIPS distance of the set conditioned on (i) the requested axis and (ii) other axes related to the prompt. When constructing our training and validation data, we ask an LLM to propose 3–5 plausible axes of variation per prompt; we use the first one for (i) and the rest for (ii). We test this across multiple diversity strengths α (the scale of the trained adapter; α = 0 is the untrained policy) and baselines. We report results for the fastest model in our setup (FLUX.2 [klein]) on 200 validation prompts:
| Method | TPIPS diversity gain (requested axis) | TPIPS diversity gain (other axes) | Pairs with more gain on requested axis |
|---|---|---|---|
| T-DIG α = 0 | 0.19 | 0.01 | 96.9% |
| T-DIG α = 0.25 | 0.24 | 0.05 | 97.4% |
| T-DIG α = 0.5 | 0.28 | 0.09 | 96.4% |
| T-DIG α = 0.75 | 0.29 | 0.11 | 97.4% |
| T-DIG α = 1 | 0.31 | 0.13 | 95.9% |
| CADS (baseline) | 0.12 | 0.12 | 48.5% |
| Group Inference (baseline) | 0.05 | 0.04 | 46.9% |
The diversity gain is larger on the requested axis than on the other axes for 96–97% of pairs at every strength. The baselines, which do not see the axis, spread their gain evenly across requested and non-requested axes, so they gain more on the requested axis for only 47–49% of pairs, i.e., chance.
Note that off-axis diversity is not expected to be zero. Axes of the same prompt can overlap in meaning: for example, if background is the requested axis and lighting is one of the other axes, changing one naturally changes the other. We also view some off-axis variation as a good thing: our goal is to show users a variety of outputs that help them explore different interpretations of a prompt, and off-axis creativity helps with that. Importantly, off-axis variation increases with diversity strength, which is a user control: at low strength there is almost no off-axis change (0.01 at α = 0), and at high strength there is some, so users can choose based on their preference.
Editing: In editing, the user starts from a set of images for a prompt and asks T-DIG to make them more diverse along one axis. T-DIG writes one edit instruction per image, and the editor applies it. We run the same test as above, but the diversity gain is measured relative to the source images (before editing) instead of an i.i.d. set.
| Strength α | TPIPS diversity gain (requested axis) | TPIPS diversity gain (other axes) | Pairs with more gain on requested axis |
|---|---|---|---|
| 0 (untrained) | 0.20 | 0.05 | 95.9% |
| 0.25 | 0.23 | 0.06 | 98.8% |
| 0.5 | 0.24 | 0.06 | 99.0% |
| 0.75 | 0.25 | 0.06 | 97.9% |
| 1 | 0.25 | 0.07 | 97.9% |
At every strength, the diversity gain is 3.5–4.2× larger on the requested axis than on the other axes, and it is larger for 96–99% of pairs. Increasing the strength from 0 to 1 adds 0.05 on the requested axis and only 0.02 on the others, so off-axis change in editing is small and nearly flat across strengths.
We will add this analysis to the paper!
Weakness 3: Separating generation from editing
We originally omitted the editing results because the findings are similar to the generation results, but for completeness we will add them to the final version of the paper (along with qualitative examples)! The results below are across different diversity strengths:
| Editor | α | TPIPS diversity ↑ | VQAScore ↑ | Drift ↓ | HPSv3 ↑ | ImageReward ↑ |
|---|---|---|---|---|---|---|
| Qwen-Image | 0 | 0.380 | 0.802 | 0.206 | 10.78 | 0.89 |
| Qwen-Image | 0.5 | 0.413 | 0.840 | 0.203 | 11.75 | 1.14 |
| Qwen-Image | 1 | 0.423 | 0.846 | 0.196 | 11.79 | 1.17 |
| SD3.5 Large | 0 | 0.272 | 0.852 | 0.206 | 11.99 | 1.20 |
| SD3.5 Large | 0.5 | 0.309 | 0.854 | 0.193 | 12.25 | 1.25 |
| SD3.5 Large | 1 | 0.323 | 0.855 | 0.195 | 12.37 | 1.28 |
| SD3.5 Medium | 0 | 0.271 | 0.847 | 0.194 | 11.57 | 1.17 |
| SD3.5 Medium | 0.5 | 0.324 | 0.856 | 0.191 | 11.91 | 1.24 |
| SD3.5 Medium | 1 | 0.349 | 0.863 | 0.209 | 11.99 | 1.24 |
| FLUX.2 [klein] | 0 | 0.363 | 0.809 | 0.178 | 11.43 | 0.97 |
| FLUX.2 [klein] | 0.5 | 0.408 | 0.833 | 0.171 | 11.91 | 1.11 |
| FLUX.2 [klein] | 1 | 0.420 | 0.840 | 0.180 | 12.19 | 1.15 |
Questions
Q1 and Q2 are addressed in our responses to the weaknesses above.
Q3: We use α = 0.5 as the default in the human studies because it balances the trade-off between diversity and alignment well. It was simply a design choice because collecting human ratings on multiple alpha values is expensive. The main automatic comparison (Table 3) reports the fully trained policy (α = 1), but since α only scales the trained adapter, the automatic metrics can be computed at any α. The off-axis and editing analyses above already report α = 0.5 alongside α = 0 and α = 1, and we will similarly add automatic diversity, alignment, and preference results at multiple α values (including α = 0.5) for the main results in the final version!
Weakness 1: Participant, pair counts (Section 4.4 vs. RQ5) and Table 1 sample sizes, W/T/L ...
Participant and pair counts (Section 4.4 vs. RQ5): Apologies for the inconsistency! Both places describe the interactive study in which we collect the test set. Section 4.4 states our planned design which was to have 80 participants contributing 5 prompts each, for 400 pairs. However, many participants who enjoyed the study contributed more than 5 prompts, so the final test set has 405 (prompt, axis) pairs from 58 contributors. The original number of submissions was 415 (the number in RQ5), but we filtered out 10 samples whose axes were judged as poor in Stage 2 (L307–309). RQ5 reports the ratings people gave to their outputs in Stage 1, and only 53 of the 58 contributors completed the exit questionnaire; so this is where the 53 / 415 numbers come from. We will update Section 4.4 in the final version!
Table 1 sample sizes, win/loss/tie counts, and sign-test definition
The numbers are correct, but they are computed while accounting for ties, which our caption did not explain. Each comparison has three possible outcomes: T-DIG wins, the baseline wins, or a tie (no majority of raters for either method). Ties are more frequent in the diversity + alignment setting, since a win there requires diversity to be greater and alignment to be greater than or equal to the baseline.
The percentage in Table 1 is T-DIG wins divided by all comparisons, ties included. A tie therefore lowers the reported win rate even though T-DIG did not lose that comparison; for example, 10 wins, 10 ties and 1 loss is reported as 10/21 = 47.6%. The sign test is two-sided and compares wins against losses only. A win rate at or below 50% therefore does not mean T-DIG lost more often than it won. We will replace the win percentages with wins / ties / losses for every cell (full table below) and state these details clearly in the text. Please let us know if we can clarify further.
Each cell is wins / ties / losses for T-DIG (α = 0.5), with the two-sided sign-test p-value.