Pith. sign in

REVIEW 4 cited by

Permutation Invariant Representations with Applications to Graph Deep Learning

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2203.07546 v1 pith:WXYKFM4W submitted 2022-03-14 math.FA cs.LG

classification math.FAcs.LG
keywords datalearningalmostarbitrarycomputationaldeepembeddingembeddings
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

This paper presents primarily two Euclidean embeddings of the quotient space generated by matrices that are identified modulo arbitrary row permutations. The original application is in deep learning on graphs where the learning task is invariant to node relabeling. Two embedding schemes are introduced, one based on sorting and the other based on algebras of multivariate polynomials. While both embeddings exhibit a computational complexity exponential in problem size, the sorting based embedding is globally bi-Lipschitz and admits a low dimensional target space. Additionally, an almost everywhere injective scheme can be implemented with minimal redundancy and low computational cost. In turn, this proves that almost any classifier can be implemented with an arbitrary small loss of performance. Numerical experiments are carried out on two data sets, a chemical compound data set (QM9) and a proteins data set (PROTEINS).

Discussion (0). Sign in to comment.

Forward citations

Cited by 4 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Monotone and Separable Set Functions: Characterizations and Neural Models

    cs.LG 2025-10 unverdicted novelty 7.0 of 10

    Exact subset-order-preserving ('MAS') embeddings need dimension ≥|V| on finite ground sets and do not exist for infinite ground sets; the paper relaxes to weakly-MAS hat-activation models with Hölder-stability and pro...

  2. Estimating the Euclidean distortion of an orbit space

    math.MG 2025-06 accept novelty 7.0 of 10

    The paper derives exact Euclidean distortion values for several orbit spaces, including cyclic quotients of C^n, seven wallpaper group quotients, and two-sided bounds for O(r), SO(r), E(r), and SE(r) actions.

  3. On the (Non) Injectivity of Piecewise Linear Janossy Pooling

    cs.LG 2025-05 accept novelty 7.0 of 10

    No piecewise linear k-ary Janossy pooling is injective on general multisets, but simple deep sets are injective on compact domains of well-separated distinct points.

  4. Optimal Transport-based Permutation-Invariant Bayesian Optimization of Offshore Wind Farm Layouts

    cs.AI 2026-03 conditional novelty 6.0 of 10

    Optimal-transport flows turn permutation-invariant layout optimization into a standard BO problem, yielding higher AEP and lower runtime than vanilla BO on a five-turbine wind-farm surrogate.

Pith tools