arXiv Machine Learning July 27, 2026 By Tasfia Nuzhat Ornee, Elias Hossain, Ivan Garibay, Niloofar Yousef

MotifRole-Diff: Risk-Optimal Role-Aware Corruption for Masked Molecular Graph Diffusion

Read the original on arXiv Machine Learning →

arXiv:2607. 21634v1 Announce Type: new Abstract: Masked discrete diffusion for molecular graph generation typically applies a uniform corruption schedule to all tokens in a lossless graph-to-sequence representation, implicitly treating structurally heterogeneous molecular components as equally difficult and equally important to reconstruct.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

diffusion

arXiv AI

6d ago

Faithful, Sufficient and Understandable: Rethinking Graph Counterfactual Explanations via Discrete Diffusion Inversion

arXiv:2608. 12083v1 Announce Type: cross Abstract: Graph Neural Networks (GNNs) achieve strong predictive performance on graph-structured data across domains such as chemistry, biology, and network analysis, yet they provide no intrinsic explanation of their predictions.

By David Bechtoldt, Sidney Bender

diffusion benchmarks safety

arXiv Machine Learning

Jun 8

Controllable Molecular Generative Foundation Models

arXiv:2605. 15354v2 Announce Type: replace Abstract: Despite the success of foundation models in language and vision, molecular graph generation still lacks a unified framework for heterogeneous design tasks with reliable controllability.

By Yihan Zhu, Yuhan Liu, Weijiang Li, Tengfei Luo, Meng Jiang

rag diffusion reinforcement-learning benchmarks

arXiv Machine Learning

Jun 2

Uncertainty-Calibrated Diffusion for Reliable 3D Molecular Graph Generation

arXiv:2606. 01595v1 Announce Type: new Abstract: Bayesian inference provides a principled framework for modeling epistemic uncertainty in neural networks by treating predictions as distributions rather than deterministic values.

By Fang Wan, Jingxiang Qu, Yi Liu

diffusion benchmarks

arXiv Machine Learning

Jun 9

Graph-GRPO: Training Graph Flow Models with Reinforcement Learning

arXiv:2603. 10395v2 Announce Type: replace Abstract: Graph generation is a fundamental task with broad applications, such as drug discovery.

By Baoheng Zhu, Deyu Bo, Delvin Ce Zhang, Xiao Wang

diffusion reinforcement-learning benchmarks

arXiv Machine Learning

Jul 14

Inverse-IMPRESSION: A Graph-based Platform for Molecular Structure Elucidation from Experimental NMR Spectroscopic Properties

arXiv:2607. 09978v1 Announce Type: cross Abstract: Here, we present a platform built on our inverted Graph Transformer Network, IMPRESSION-G2, which can accurately and rapidly reconstruct molecular bonding directly from experimental nuclear magnetic resonance (NMR) spectroscopic information.

By Zheqi Jin, Grace Armitage, Richard Cox, Ben Honor\'e, Mohammad Golbabaee, Craig Butts

llms

arXiv AI

Jun 2

Graph Energy Matching: Transport-Aligned Energy-Based Modeling for Graph Generation

arXiv:2603. 23398v3 Announce Type: replace-cross Abstract: Generative modeling of discrete data, such as graphs, underpins many scientific and industrial applications, including molecular discovery and materials design.

By Michal Balcerak, Suprosanna Shit, Chinmay Prabhakar, Sebastian Kaltenbach, Michael S. Albergo, Yilun Du, Bjoern Menze

diffusion benchmarks

Related stories

Faithful, Sufficient and Understandable: Rethinking Graph Counterfactual Explanations via Discrete Diffusion Inversion

Controllable Molecular Generative Foundation Models

Uncertainty-Calibrated Diffusion for Reliable 3D Molecular Graph Generation

Graph-GRPO: Training Graph Flow Models with Reinforcement Learning

Inverse-IMPRESSION: A Graph-based Platform for Molecular Structure Elucidation from Experimental NMR Spectroscopic Properties

Graph Energy Matching: Transport-Aligned Energy-Based Modeling for Graph Generation