- Rating
- 2
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This work presented solver-sufficient symbolization, where commitments reduce the optimization problem to one whose solution is simplified by discrete commitments and relative affine-TU certificates, which allow LP relaxation in exact manner after conditioning the problem on symbolic assignments. In the case of neural energy objectives, τ-descent faithfulness, which is an accessibility requirement relying on energy increase bounds, is introduced.
Critical strengths
Relative affine-TU constructions, in the context of PWA, provide a way to verify sufficient conditions for LP relaxation to be exact after symbolic assignment. This matters since it provides a concrete target for symbol discovery.
Critical weaknesses
Please check Questions For Authors
Questions for authors
Definition 3.2 defines v^tau{s,R}(x) via an infimum taken over all possible admissible continuous paths. Thus, Proposition B.4 ensures that the correct endpoint is reachable according to the path framework. However, it does not guarantee that one single execution of a gradient descent or noisy descent algorithm achieves this target and how many steps are necessary for this.
Appendix D assigns the maximum value of the objective function at three points (endpoints and middle point) for each edge and suggests that in this way the barrier estimation for the segment is not underestimated. This statement is false, since the maximum value at three sampled points might be lower than the maximal value of the function on the whole segment. The graph will miss all barriers.
What are the neural energy functions and how do they convert the baselines? The authors need to clarify hs(x), describe how it works, discuss the management of the diffusion time and the modification of each baseline.
Experiments do not provide the evidence of the optimality preservation and even lead to the degradation of the quality. In the case of neural networks, Table 2 shows Aligning success deteriorating from 0.94 to 0.84 if symbols are added.
6 comments
Response to Reviewer 9u1x (Part II: Q2, Q3)
Q2 (The maximum over three points can underestimate the barrier along a segment, so the statement in Appendix D is false.)
The reviewer is right; we thank the reviewer for catching this writing error and have corrected it. For a neural energy, the exact maximum along a segment has no closed form, so sampling gives a provable lower bound on it. In our implementation for all results in the paper, we evaluate 50–200 points per segment depending on the task, not three (cf. Fig. 4).
Q3 (What is the neural energy , how is the diffusion time handled, and how is each baseline converted?)
We clarify below how is constructed for each baseline and how the diffusion time is handled; the construction for each baseline is listed in App. F.2 of the revised version. In short, is up to a constant, or a surrogate of it when the likelihood has no closed form; it is evaluated on a fixed grid of diffusion times with fixed noise; and every baseline is frozen.
Energy. One rule covers all policies: viewing a trained policy as a conditional distribution over the decision , we set , where is independent of and affects neither minimizers, gradients, nor barrier heights. When the likelihood has no closed form (e.g., diffusion policies), we use a surrogate of . For DDPM, this is the denoising training loss at , with fixed pairs of diffusion time and noise and the correspondingly noised :
This is the DDPM variational bound (ELBO) without its time weights on a fixed set of diffusion times: a surrogate of , not a bound on it. BESO, also a diffusion policy, uses the same construction with its own denoising loss. Explicit energy models (IBC) give , and explicit densities (e.g., GMM) give directly. For most baselines (DDPM, BESO, IBC), is thus read out from the frozen network with no additional training; for ACT and BeT, it is an energy network fitted to action samples from the frozen policy. Values come from forward passes and gradients from automatic differentiation with respect to .
Diffusion time. Each sets a smoothing scale of the energy: at small the energy is sharp and has many spurious basins (not corresponding to distinct action modes); at large it is over-smoothed and distinct modes merge. We therefore use a fixed grid of intermediate with fixed noise (four diffusion times combined with four fixed noise draws, pairs, uniform weights), which makes deterministic and its basin structure stable.
Baselines. No baseline network is modified: each is frozen and its energy is obtained as above, with the same solver and budget for all. The energy of each baseline is listed in App. F.2.
Response to Reviewer 9u1x (Part III: Q4)
Q4 (The experiments give no evidence that optimality is preserved, and quality even degrades on Aligning (0.94 to 0.84).)
Thanks. Here, we add direct evidence that optimality is preserved, and we explain the drop on Aligning. (i) In the PWA setting, the conditioned LP attains the MILP optimum on all tested states; in the neural setting, the symbol-guided solver ends within of the best value found by multi-start on 91% (Avoiding) and 100% (Pushing) of the states. (ii) The drop on Aligning comes from the descent step limit and from under-sampling; once both are addressed, SR returns to 0.94. We also respectfully note that the paper already reports optimality evidence through (Table 2).
Guarantee and evidence. Reaching the optimum requires (a) that the optimum of each conditioned problem is reachable and (b) that the solver reaches it. Our properties guarantee (a); (b) depends on the given solver (Q1).
- PWA setting. Total unimodularity makes the LP relaxation exact (a) and LP solvers are exact (b), so each conditioned problem is solved to optimality without further assumption. Empirically, the conditioned LP at , with the MILP optimum, attains the MILP optimal value on all tested states.
- Neural setting. If a region has the -descent faithfulness property (verified on samples), a path to a near-optimal point exists from every start in it (a); whether the solver follows it to the end (b) depends on the solver, e.g., on its step limit. measures this reachability: the minimal number of descents needed to reach the -neighborhood of the optimum with probability at least (Table 2). In addition, the symbol-guided solver reaches this neighborhood, relative to the multi-start optimum, on most states.
| States where the optimum is reached (%) | MPushT | Hook | Shelf | Block | Cuboid | Pen | Avoiding | Pushing |
|---|---|---|---|---|---|---|---|---|
| exact (PWA) / within (neural) | 100 | 100 | 100 | 100 | 100 | 100 | 91 | 100 |
The drop on Aligning. Its cause is consistent with the decomposition above: the step limit violates (b), and under-sampling leaves a few uncertified regions that violate (a). A descent that occasionally falls into such a basin needs more steps to leave it, so we allow more descent steps; denser sampling (1.2× the samples) then removes these regions. Together they restore SR to that of the multi-start solver, while the advantage remains.
| Aligning | DDPM (multi-start, 24 starts) | DDPM (Symbols, Table 2; ≈2.5 starts) | + more descent steps | + more descent steps + denser sampling (1.2×) |
|---|---|---|---|---|
| SR | 0.94 | 0.84 | 0.88 | 0.94 |
| 10.31 | 2.41 | 2.12 | 2.36 |
Exact certification is out of reach for modern networks; our framework instead makes its premises explicit and gives testable diagnoses when they fail.
We have incorporated these corrections, clarifications and results into the revised version. We hope they address your concerns; if so, we would be grateful if you would consider raising your score.
Response to Reviewer 9u1x (Part I: Q1)
We thank the reviewer for the precise technical comments, which help us correct and sharpen the paper. We address each point below.
Q1 (Prop. B.4 shows that a near-optimal endpoint is reachable, but not that a single descent reaches it, nor how many steps this takes.)
Thanks for bringing this up. We respectfully note that, in our setting, the solver and its capability are given: the paper studies how to partition a problem, not how to solve each part, and the solver is chosen by the user (Eq. 2 is defined with respect to a given solver). We agree that Prop. B.4 shows only that a good path exists, not that any given optimizer would follow it. Here, we support the practicality of Prop. B.4 empirically and with existing theory, as detailed below.
(1) Empirically. Our implemented solver, projected noisy gradient descent (a Langevin-type dynamics), reaches the τ-neighborhood in most cases: with one confined descent per region, its best endpoint is within τ of the best value found by a 200-start multi-start on 29/32 states (Avoiding) and 30/30 states (Pushing); see Q4.
(2) Theoretically. There are established methods that provably ensure that an optimizer that can climb by τ is not trapped in basins shallower than τ, given enough steps; for such an optimizer, a τ-descent faithful region has no basin that keeps it away from the τ-neighborhood of the optimum. We give two examples. For simulated annealing, Hajek [1] defines the depth of a local minimum as the smallest climb needed to reach a lower state, which is the quantity that τ bounds, and proves (for finite state spaces) that annealing leaves every local minimum whose depth is at most its cooling constant. Projected Langevin algorithms [2], which have the same form of update as our solver, provably reach near-optimal points of non-convex objectives on compact convex sets. These methods act on the optimizer side, i.e., on how to solve each part, which our method takes as given and does not modify.
Number of steps. The number of steps is likewise a property of the given solver, so the paper gives no bound on it. Determining it is an inherent challenge of diffusion- and flow-based energy methods; our method inherits it and does not modify it. The simulated tasks (Table 2) use 600 steps per descent.
In the revised version, we have stated the solver assumption (see Remark 3.2, to which Prop. B.4 now refers), added the two points above, and rewritten the three sentences that read as a guarantee for the implemented descent (Sec. 3.1 after Eq. (5), App. A.1, and App. B.5). In the PWA setting, no such assumption is needed, since both the certificate and the LP solver are exact. Thank you very much again for raising this concern.
[1] Cooling Schedules for Optimal Annealing, 1988.
[2] Projected Stochastic Gradient Langevin Algorithms for Constrained Sampling and Non-Convex Learning, 2021.