- Rating
- 3(+1)
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 3(+1)
Read review
Paper summary
This paper studies the token ordering problem in point-cloud Mamba models. Existing approaches typically rely on fixed geometric orders, which cannot adapt to the task during training. To address this limitation, the authors propose GCSO, which introduces trainable local soft routing around a geometric base order while keeping token participation balanced. Experiments on point-cloud classification, segmentation, and few-shot learning show consistent improvements over existing Mamba-based baselines, with particularly strong results on ScanObjectNN.
Critical strengths
- The motivation is clear and easy to follow. The paper identifies a meaningful limitation of fixed geometric ordering in point-cloud SSMs and provides a straightforward way to make the routing trainable.
- The experimental results are strong, especially on ScanObjectNN. GCSO consistently outperforms several recent Mamba-based baselines, and the ablation studies show that the proposed components contribute to the final gains.
Critical weaknesses
The “soft ordering” interpretation is not fully convincing. GCSO does not actually reorder the original tokens into a new sequence. Instead, it computes , where each output slot is a weighted mixture of several neighboring tokens, which looks more like an adaptive token-mixing layer than an ordering operation. The paper therefore needs to clarify whether the gain really comes from learning a better order, or simply from adding this extra feature mixing before Mamba.
Questions for authors
- Could the authors clarify in what sense GCSO actually performs “ordering”? Since each output slot mixes several neighboring tokens, it seems closer to a token mixer than an actual reordering operation.
- If GCSO does learn a meaningful new order, a visualization of the original and learned orders would help demonstrate this. Otherwise, a comparison with a similar local token-mixing baseline would be useful.
- Could the authors show what kind of ordering GCSO actually learns, either quantitatively or with a simple visualization?
My score is currently borderline, but I would consider raising it to 3 if these concerns are addressed.
7 comments
Response to 36vZ-W1 (Part 2/2)
The Birkhoff–von Neumann theorem guarantees that a doubly stochastic can be expressed as a convex combination of hard permutation matrices. Here the off-diagonal entries give the coefficient of , and give that of ; the remaining coefficient is for . Thus, at the scan-input level, our soft ordering is a linear combination of hard orderings: . Each hard matrix selects one source per slot; for example, makes Mamba read . Their weighted combination instead gives the three mixed features above. Relative to the original order, retains 70% of its reading weight at slot 1 and contributes 30% to slot 2; contributes 30%, 40%, and 30% to slots 1, 2, and 3; and contributes 30% to slot 2 and 70% to slot 3. Thus Mamba's second slot reads all three nearby sources rather than only . Hard ordering exchanges whole tokens; the soft assignment changes how each source contributes to the ordered input.
This continuous formulation has two direct benefits. First, a binary hard assignment cannot be adjusted smoothly by ordinary task gradients; gradients can instead pass through to the local offset predictor, changing Gaussian affinities and fractional reading weights. Second, retains both participation budgets of a hard permutation: every slot reads one unit of feature weight and every source contributes one unit in total. This avoids excessive repeated reading or complete omission of a source while the scan input changes. Soft ordering can also change non-dominant weights without swapping the largest-weight source of any slot: in the example above, the dominant sources at slots 1, 2, and 3 remain , yet the fractional weights make Mamba read a feature sequence different from . A new hard permutation is therefore not required for soft ordering.
The forward soft assignment is both a continuous allocation to ordered scan slots and an adaptive local token-mixing operation; in GCSO these describe the same computation. We further compare it with fixed local mixing under the same routing interface. On PB_T50_RS/ModelNet40, full GCSO reaches 94.20%/95.70%, compared with 93.27%/95.30% for fixed Gaussian mixing and 93.58%/95.02% for frozen offsets. These comparisons support the benefit of learned, input-dependent soft assignment over fixed mixing in the tested architecture.
The three-slot example is on p. 16, and the fixed-mixing controls are in Table 6 on p. 18 of the revised manuscript.
Response to 36vZ-Q1, Q2, and Q3
36vZ-Q1
We appreciate this constructive comment. We define soft ordering as a continuous assignment of source tokens to ordered scan slots, and W1's three-slot example shows it is a linear combination of hard orderings at the scan-input level. We use “ordering” in the sense of an assignment matrix from source tokens to ordered scan slots. A hard order is a binary permutation matrix : each source receives one integer slot, and each slot receives one source. GCSO relaxes the binary entries to a nonnegative continuous matrix while retaining . Column of therefore assigns source fractional participation in nearby scan ranks rather than one unique hard rank; row specifies the source weights read at rank . The doubly stochastic constraint gives the mathematical link to hard ordering: lies in the convex hull of hard permutation matrices and retains the hard order's two unit assignment budgets. Learned local sequence offsets make this soft-order assignment responsive to the current point cloud.
As derived in W1, consider a Morton base sequence with features . The illustrative shows soft ordering as a linear combination of three hard orders: 0.4 times the original order, 0.3 times a swap of positions 1 and 2, and 0.3 times a swap of positions 2 and 3. Reading its columns, participates with weights 70% and 30% at ranks 1 and 2; has weights 30%, 40%, and 30% across ranks 1, 2, and 3; has weights 30% and 70% at ranks 2 and 3. A hard order places a whole source at an integer rank; this soft order distributes its participation continuously across nearby ranks.
The ordering still has to supply a sequence model: reading the rows of , Mamba receives , , and at slots 1, 2, and 3, and propagates state along those ordered slots. Even if the largest-weight sources remain , soft ordering changes their participation across scan ranks and changes Mamba's ordered input. Thus “token mixer” describes the forward computation, while “soft ordering” describes the constrained continuous assignment it implements; both refer to the same operation. The fixed-Gaussian and frozen-offset controls in W1 test whether learning this assignment adds value beyond fixed local mixing.
The corresponding three-slot illustration appears on p. 16 of the revised manuscript.
36vZ-Q2
We provide a visualization of the original scan and the learned GCSO reading targets in the revised manuscript. Figure 4 shows all 128 slots of the first Morton path in layer 12 on a real ModelNet40 test airplane: the left panel connects the original patch centers in the original scan order, and the right panel (GCSO) connects the interpolated targets derived from learned sequence-index offsets. At the scale of the whole object, individual local changes are hard to trace. Figure 5 therefore shows three disjoint ten-slot windows from early, middle, and late parts of the same scan: A (13–22), B (76–85), and C (109–118). In each local comparison, gray dots are sampled airplane points; colored circles in the left panel are original patch centers , and those in the right panel, titled GCSO, are interpolated targets . The panels share a projection, point-cloud background, and axis limits. Solid lines join consecutive slots within each window, while the faint dashed line in the GCSO panel shows the original local path. In A, a long excursion between original centers is drawn into a more compact group of targets; in B, the lower branch shifts inward while a longer connection remains; in C, the selected zigzag becomes narrower.
The corresponding full-path and local visualizations appear as Figures 4 and 5 on p. 18 of the revised manuscript.
36vZ-Q3
We provide the whole-path context in Figure 4, the three local comparisons in Figure 5, and a quantitative path-length measurement across all 581 OBJ_ONLY test shapes. The selected A, B, and C windows show distinct local changes in the candidate reading targets. As separate quantitative evidence, we measured the final weighted-center paths across all 581 OBJ_ONLY test shapes: the mean ratio of successive-center distance after versus before routing is 0.906 for the first Morton path across 12 layers and 0.903 for the complementary path in layers 4, 8, and 12. Thus the routed paths retain 90.6% and 90.3% of their original lengths, respectively. Together, the visualization and path-length ratios characterize local, position-dependent soft assignment with moderate geometric smoothing.
The visualizations appear as Figures 4 and 5 on p. 18 of the revised manuscript.
Finally, we sincerely thank the reviewer for the time and effort devoted to evaluating our work.
Response to 36vZ-W1 (Part 1/2)
Thank you for pointing out that H=PΠH acts as an adaptive token-mixing layer. We agree: token mixing describes the forward computation, while soft ordering describes the constrained assignment of source features to ordered scan slots. Below we explain this assignment and use fixed-mixing controls to assess whether the gain goes beyond extra feature mixing before Mamba. To make the terminology self-contained, H is the patch-feature matrix in original patch-index order, with each row representing one patch token; Π arranges these tokens by a geometric Morton order; and P determines how sources in that base sequence contribute to the scan slots Mamba reads in order. We call this “soft ordering” because the essential effect of an order for Mamba is not to output a list of indices, but to determine which features occupy its successive input positions.
Consider first a hard order. Let A be a binary permutation matrix: Aij=1 means scan slot i reads source token j in the geometric base sequence, and each row and column has exactly one 1. The hard assignment selects one source per slot, and AΠH then presents the slot-ordered feature sequence to Mamba. For this illustration, label the three rows of ΠH as h1,h2,h3. If A swaps the first two positions, the operation is
AΠH=010100001h1\h2\h3=h2\h1\h3.
Mamba then reads h2,h1,h3 at slots 1, 2, and 3. Assigning tokens to positions and forming a feature sequence in those positions are thus two linked parts of hard ordering.
GCSO makes both parts continuous. Current features and geometry predict local sequence-index offsets, which produce Gaussian reading affinities between scan slots and nearby source tokens; these affinities are then turned into a nonnegative conservative matrix P. Here Pij is the weight with which source j enters slot i, and Hi=∑jPij(ΠH)j is the feature Mamba actually reads at that slot. Alongside the hard swap above, consider this illustrative soft assignment:
PΠH=0.70.300.30.40.300.30.7h1\h2\h3=0.7h1+0.3h20.3h1+0.4h2+0.3h30.3h2+0.7h3.
The same P can be derived from three named hard orders. Let I keep the original sequence, S12=010100001 swap positions 1 and 2, and S23=100001010 swap positions 2 and 3. Entrywise addition then gives
0.4I+0.3S12+0.3S23=0.4+0.30.300.30.40.300.30.4+0.3=P.