- Rating
- 1
- Confidence
- 4
- Informativeness
- 2
- Presentation
- 1
- Soundness
- 1
Read review
Paper summary
The paper extends the divide-and-conquer idea of AI Feynman and proposes a more comprehensive framework for identifying reducible substructures in a complex data-generating function, thereby facilitating easier downstream SR tasks. The irreducible substructures are then fed to existing SR methods for symbolic search, and the final expression is derived by combining the symbolic expression of each irreducible substructure.
Critical strengths
- The paper describes a comprehensive list of potential reducible structures of the data-generating function .
- The empirical study covers a comprehensive list of datasets.
Critical weaknesses
- The authors mentioned DDRSR can be paired with "different downstream SR methods," but their evaluations in Section 4.1 only focused on quite dated SR methods published in 2021 to 2023. In a 2026 paper, readers will expect to see much more recent methods. Furthermore, the authors hand-pick baseline SR methods in subsections of 4, rather than reporting the effect of DDRSR on all 5 baseline SR methods. For instance, Table 1 only considers 3/5 SR baselines, Table 3 only considers 1/5 SR baselines, etc. More importantly, Table 4 shows that the downstream solvers receive only 40 samples in , which explains the negative baseline at 6+ variables in Table 7 and inflates the apparent benefit of decomposition, and Table 1 shows DDRSR uses 2–7$\times$ the runtime of the baseline with no compute-matched comparison.
- The evaluated setting is not symbolic regression from data. Every criterion queries at chosen points: slices with variable groups fixed to constants, two-point rectangles for the constant-offset recovery, and finite differences at step (Table 4: 20 query points 10 constant settings). Appendix C concedes that the method needs "an active sampling environment, such as a directly queryable physical field". AI Feynman handles this with a neural surrogate, which is never used here, so neither the actual AI Feynman system nor a data-only DDRSR is ever run, and the abstract and introduction present the results as benchmark gains without this caveat.
- The uniform noise added to the data is insignificant compared to the signal: the largest amplitude, on inputs in , equals the finite-difference step in Table 4 and is three to four orders of magnitude below the noise levels used in SRBench. A proper signal-to-noise ratio control is needed. See Ye & Li (arXiv:2410.13681; ICML 2025) for instance.
- The paper needs a significant rewrite so its presentation is readable to human audiences.
- Notation issues: seems to imply is infinitely dimensional, but it is actually -dimensional. More critically, Section 3.2.1 uses without defining and . The operation of is also very confusing, as it implies matrix multiplication, but 's are not matrices and are definitely of arbitrary dimension, yet is not differentiated, so makes no sense. As far as I can tell, seems to be a subvector of with arbitrarily selected dimensions . If so, why not use , where ?
- The paper also did not mention what datasets were used in the empirical study. Figures 2(c)-(d) suggest there are multiple datasets used in the empirical study, but none are discussed or cited in the main paper. The figures also suggest a set of datasets called "Others", which Table 5 in Appendix F never labels.
- Section 4.2 mentioned that only the "best-performing baseline SR methods" are reported in the comprehensive SR experiments, but how is performance measured? You seem to have results for all 5 SR methods. Why not just report all of them like in Figures 2(e)-(f)? Same thing for Section 4.4. Why not report the runtime for all SR methods? Furthermore, the Appendices list the CPU and GPU used for the empirical study, but do not report which methods run on CPU and which run on GPU, so the reported runtime is not useful.
- The list of writing issues just doesn't stop.
- The "Relative Separation Score" is also oddly defined in Section 4.1, where all mid-field methods are assigned a score of 0.5, leaving nothing to differentiate them. Furthermore, the best method is given a score of 1, but there is no telling how many decomposition patterns it misses. It is also defined relative to the competing method, so DDRSR, whose rule set is a superset of AI Feynman's, scores 1 by construction, and the "maximum number of correct decomposition patterns" is never operationalized.
Questions for authors
See weaknesses above. Also,
- Section 4.3 states that the custom complex-test datasets are excluded from top-down expression separation and are used solely for bottom-up variable composition. What's the justification?
- Can DDRSR run from a fixed dataset, e.g., through a neural surrogate as AI Feynman does? Please report recovery rates under that setting.
- Does the advantage survive noise amplitudes of to ?
- Why disable AI Feynman 2.0's candidate search for , given that it is AI Feynman's mechanism for the forms you claim it cannot handle? Please run the published AI Feynman 2.0.
- Which 23 expressions fail to decompose at zero noise, and were any accepted decompositions wrong?
- The decomposition criteria are necessary conditions checked at 10 random points with a relative threshold of 0.01 (Appendix B), and their false-positive rate is never measured. Section 4.5 nonetheless asserts that false positives are "unlikely", and the claim that DDRSR "is generally expected to perform no worse than direct symbolic regression" rests on it.
Thanks to the authors for their detailed and careful rebuttal. Some of my concerns have been addressed, and the quality of the paper has improved. After re-examining the paper, I believe my score is already reasonably appropriate, so I plan to keep my score.