arXiv Machine Learning July 30, 2026 By Xin Xu, Siru Tao, Kaizhen Tan

Automorphism-Induced Non-Canonicity in Top-k Explanations of Graph Neural Networks

Read the original on arXiv Machine Learning →

arXiv:2607. 26344v1 Announce Type: new Abstract: A gradient-based GNN explainer given a molecule with two chemically equivalent nitro groups assigns them attribution scores that are equal to the last bit.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

efficiency

arXiv AI

Aug 12

Reasoning Shortcuts and Value Symmetries: What Symmetry Permits, Architecture Realizes, and Optimization Selects

arXiv:2608. 10420v1 Announce Type: new Abstract: Reasoning shortcuts are solutions of a neurosymbolic system's rules that produce correct predictions through unintended concepts.

By Xin Xu

rag benchmarks

arXiv Machine Learning

Jul 7

Observable- and Positional-Encoding-Dependent Symmetry Readout from Neural Network Weights

arXiv:2607. 03108v1 Announce Type: new Abstract: Post-hoc analysis of trained neural network weights often seeks to recover geometric structure directly from the parameters.

By Naoya Chiba, Satoshi Sugiyama, Yuki Uranishi

arXiv Machine Learning

Aug 10

Equivariant Sparse Autoencoders: Mechanistic Interpretability of Neural Networks on Symmetric Data

arXiv:2511. 09432v2 Announce Type: replace Abstract: Machine learning (ML) models achieve remarkable performance but remain hard to interpret due to their scale and complexity.

By Ege Erdogan, Ana Lucic

safety

arXiv AI

Jun 15

When the Tool Decides: LLM Agents Defer Blindly to Graph Neural Network Tools, and Stronger Backbones Defer More

arXiv:2606. 14476v1 Announce Type: new Abstract: A growing line of work equips large language model (LLM) agents with graph neural networks (GNNs) as callable tools, assuming the agent exercises judgment over when and how much to rely on such a tool.

By Zhongyuan Wang, Pratyusha Vemuri

llms agents

arXiv Machine Learning

Jun 30

Lost in Aggregation: On a Fundamental Expressivity Limit of Message-Passing Graph Neural Networks

arXiv:2603. 14846v3 Announce Type: replace Abstract: We define an information-complexity property for aggregation functions, capturing a vast range of practical aggregations, and prove that any Message-Passing Graph Neural Network (MP-GNN) model with such aggregations induces only a polynomial number of equivalence classes on all graphs - while the number of non-isomorphic graphs is super-exponential (in number of vertices).

By Eran Rosenbluth

arXiv Machine Learning

Aug 13

AutoGrable: What Is a Good Graph for a Table?

arXiv:2608. 11431v1 Announce Type: new Abstract: Graph learning presupposes a graph, and tables and relational databases do not come with one.

By Tamara Cucumides, Floris Geerts

safety

One email a morning, machine-written

Related stories

Reasoning Shortcuts and Value Symmetries: What Symmetry Permits, Architecture Realizes, and Optimization Selects

Observable- and Positional-Encoding-Dependent Symmetry Readout from Neural Network Weights

Equivariant Sparse Autoencoders: Mechanistic Interpretability of Neural Networks on Symmetric Data

When the Tool Decides: LLM Agents Defer Blindly to Graph Neural Network Tools, and Stronger Backbones Defer More

Lost in Aggregation: On a Fundamental Expressivity Limit of Message-Passing Graph Neural Networks

AutoGrable: What Is a Good Graph for a Table?