- Rating
- 3
- Confidence
- 2
- Informativeness
- 3
- Presentation
- 2
- Soundness
- 3
Read review
Paper summary
The paper proposes FMTR, a batch black-box optimization method that learns a flow model for proposal distribution, conditioned on previous observations and remaining budget. The flow model is pretrained with data that covers many generated synthetic objectives and trajectories, on a value-weighted target distribution derived from Thompson sampling, and is reusable at inference time, conditioned in-context on observations. FMTR is shown to generally outperform many batch black-box optimization methods across synthetic and real-world functions. Ablations on various design decisions are also included.
Critical strengths
- The work is novel, pretraining a single flow model for an in-context conditional proposal distribution (across multiple dimensions and batch sizes), compared to online fitting which existing (generative-based) works do. Many design decisions key to the pretraining, such as the construction of synthetic objectives and tilted target, are also detailed and supported by ablation studies.
- Good empirical results for FMTR. Best final average ranks and faster wall-clock time than existing batch BBO methods.
Critical weaknesses
- The reason for FMTR (without Langevin and tilt) outperforming qTS is still unclear, given that TS targets were used to train the flow model. I wonder if an analysis of the different priors used by each method could be informative. Especially since there seems to be a center bias for synthetic objectives generated for pretraining (Section B.2, center for sampled from with ).
Questions for authors
Clarifications:
- For Eq. 10, what is , in , , and ? Temperature? Can they be introduced briefly in words, when first used?
- Have the authors tried a constant (and constant )? For comparison with the scheduled to see how much the round-awareness matters.
- FMTR implicitly represents the prior in the flow model to approximate the posterior distribution (if I'm understanding correctly). Is it possible to observe this prior to compare with qTS's prior, through the flow model without conditions or via pretraining data? Also, what is the prior used for qTS?
4 comments
Rebuttal 2/2
Q4: The reason for FMTR (without Langevin and tilt) outperforming qTS is still unclear, given that TS targets were used to train the flow model. I wonder if an analysis of the different priors used by each method could be informative.
We agree that the difference lies in the priors. Without refinement and tilt, FMTR amortizes Thompson sampling under its pretraining prior, a mixture of kernel families and randomized hyperparameters with a global trend component. During pretraining, the TS targets are constructed using the known generating kernel and hyperparameters of each synthetic task. Conversely, qTS performs Thompson sampling under a stationary GP whose hyperparameters are estimated online from few observations. The two are therefore TS under different priors, and their relative performance should depend on which prior better matches the test functions.
To test this, we evaluated both methods on functions drawn from qTS’s own GP prior (RBF kernel, lengthscale , no trend). Since the exact value of the sampled lengthscale leading to a function draw to be optimized can be retrieved, we gave it to the GP surrogate powering qTS, leaving it frozen during the trial. In this case, qTS is perfectly well-specified. On the contrary, such function draws occupy only a small part of FMTR’s broader pretraining prior (FMTR's lengthscale range is , and includes trends). In this setting, qTS outperforms the TS-only FMTR variant (See Figure S5, Appendix A.5).
These results suggest that FMTR's advantage on our original benchmarks stems primarily from differences in prior specification, rather than from amortization itself. FMTR is pretrained on a diverse mixture of kernels, hyperparameters, and global trends, whereas qTS relies on a stationary GP with hyperparameters estimated from limited observations. When the GP prior is correctly specified, qTS outperforms the TS-only FMTR variant, as expected. Needless to say that this is seldom the case, and in the limited observations regime considered in this submission, learning hyperparameters seems wasteful. We have added this experiment and discussion in Appendix A.5.
Q5: Have the authors tried a constant (and constant )? For comparison with the scheduled to see how much the round-awareness matters.
We thank the reviewer for this excellent suggestion. We trained a variant (scheduler D) with constant temperature , and hence a constant refinement budget in every round. Since its target no longer depends on the remaining horizon, we also removed from the model input. We chose because it equals the mean temperature of scheduler A over the horizon, so this variant isolates the effect of the schedule's shape at a matched exploitation level. The result is shown in Figure S11 (Appendix A.8).
Averaged over dimensions, scheduler D outperforms schedulers B and C but remains behind scheduler A. The gap between A and D grows with dimension: D performs on par with A for , whereas A clearly outperforms it for . We draw two conclusions.
First, even without explicit horizon conditioning, FMTR's target adapts over the rounds: as observations accumulate, the posterior contracts and the target concentrates on high-value regions. This explains why a constant temperature is already competitive, especially in low dimensions, where this contraction is fast.
Second, explicit round-awareness provides an additional gain when the schedule is appropriate (scheduler A), particularly in higher dimensions, where the history alone constrains the posterior less. At the same time, the other schedules (B, C) do not outperform the constant temperature, showing that performance depends on how exploitation is distributed across the rounds, not only on whether the target depends on the horizon. The main gains come from refinement and value weighting (Fig. 4b); the temperature schedule is a secondary but beneficial component whose effect grows with dimension.
Thank you for the detailed response. I appreciate the additional experiments on shifted synthetic functions and constant temperature, the further analysis of FMTR without refinement and tilt, and the prior visualizations. I will maintain my positive score.
One question on the prior visualization (Figure S15): how were the argmax points computed? Were they from a discrete candidate set? Visually, the two subplots seem to have several points at matching spots.
Thank you for the response
Dear Reviewer JaW2,
Thank you for your response. We sincerely appreciate your time and consideration.
Regarding your question about the prior visualization, we computed the argmax locations by evaluating each sampled function at 3,000 points and selecting the points with the highest value. Although this approach uses a discrete set of points, the sampling density is sufficient for the two-dimensional domain . We will clarify this procedure in the revised manuscript.
Best regards,
The authors
Rebuttal 1/2
We thank the reviewer for their thoughtful feedback on our methodology, which helped us clarify key design choices and strengthen the manuscript.
As derived in Appendix D.3 of the revised manuscript, for a fixed pair (D,α), FMTR is trained to approximate ptarget(x∣D,α)=Ep(τ,c∣D,α)[p~1(x∣τ,D,c)], where τ denotes the synthetic task, c=(h,H), and p~1 is the tilted target in Equation 13. The pretraining prior enters through p(τ,c∣D,α). FMTR therefore implicitly represents its prior, but its target is a task-averaged refined proposal rather than a posterior over the maximizer.
Regarding qTS's prior, qTS uses the default BoTorch
SingleTaskGP: an RBF kernel with lengthscale hyperprior ℓi∼LN(2+21logd, 3), a noise-variance hyperprior LN(−4,1), and a constant mean on standardized outputs.Finally, FMTR cannot be conditioned on an empty history, because its decoder cross-attends to the context embedding (Figure S8). We therefore compare the priors through the distribution of maximizer locations they induce. For each prior, we draw 1000 independent 2D functions and record each maximizer; for qTS, lengthscales are sampled from the hyperprior above. Figure S15 (Appendix A.11) shows that both distributions spread over the whole search space. Thus FMTR's pretraining prior does not show a pronounced center bias. Trend centers are drawn from U(−0.375,0.375) only with probability 0.75 and from U(−0.8,0.8) otherwise, and the GP component further displaces the maximizer.
To address this follow up question, we evaluate on shifted versions of Ackley, Levy and Rosenbrock. We keep each objective unchanged and shrink the search box so that the global minimizer x⋆ is no longer near the center:
Ackleyshiftuses [−32.768,2]d (x⋆=0),Levyshiftuses [−10,2]d (x⋆=1), andRosenbrockshiftuses [−5,2]d (x⋆=1). In the normalized [−1,1]d box that FMTR sees, the optimum lies at about +0.88, +0.83 and +0.71 per coordinate, well outside the central region favored by the prior. The result is shown in Figure S6. We can clearly see that FMTR still achieves good performance in this scenario.T(h,H) is a temperature controlling how strongly the target favors exploitation: it sets both the tilt strength in Equation 13 (lower T concentrates mass on high-value regions) and the number of refinement steps Kh,H in Equation 10. It decreases linearly from Tmax=1.0 in the first round to Tmin=0.4 in the last. We added a brief introduction of these parameters in the revised manuscript.