Benchmarking Compositional Generalisation for Machine Learning Interatomic Potentials

• URL: http://arxiv.org/abs/2605.08988v1
• Date: Sat, 09 May 2026 15:12:59 GMT
• Title: Benchmarking Compositional Generalisation for Machine Learning Interatomic Potentials
• Abstract summary: We propose a benchmark consisting of four tasks that require some form of compositional generalisation.<n>In each task, models are tested on molecules that were unseen during training, but the training data is chosen such that generalisation to the test examples should be feasible.<n>Our empirical analysis shows that the considered tasks are highly challenging for state-of-the-art models.
• Score: 17.46673779036315
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Machine Learning Interatomic Potentials play a fundamental role in computational chemistry and materials science, enabling applications from molecular dynamics simulations to drug design and materials discovery. While recent approaches can estimate inter-atomic forces with high precision, it remains unclear to what extent they can generalise to previously unseen molecules. Do they learn the compositional structure of chemistry, capturing how molecular fragments and their combinations determine properties, or do they primarily learn to interpolate patterns that are specific to the training examples? To address this question, we propose a benchmark consisting of four tasks that require some form of compositional generalisation. In each task, models are tested on molecules that were unseen during training, but the training data is chosen such that generalisation to the test examples should be feasible for models that learn the underlying physical principles. Our empirical analysis shows that the considered tasks are highly challenging for state-of-the-art models, with errors on out-of-distribution examples often an order of magnitude higher than on in-distribution examples, even when using foundation models that have been pre-trained on millions of molecules.

Related papers

• Zatom-1: A Multimodal Flow Foundation Model for 3D Molecules and Materials [51.342983349686556]
  General-purpose 3D chemical modeling encompasses molecules and materials, requiring both generative and predictive capabilities.<n>We introduce Zatom-1, the first end-to-end, fully open-source foundation model that unifies generative and predictive learning of 3D molecules and materials.
  arXiv  Detail & Related papers  (2026-02-24T20:52:39Z)
• Foundation Models for Discovery and Exploration in Chemical Space [57.97784111110166]
  MIST is a family of molecular foundation models trained on large unlabeled datasets.<n>We demonstrate the ability of these models to solve real-world problems across chemical space.
  arXiv  Detail & Related papers  (2025-10-20T17:56:01Z)
• Predicting ionic conductivity in solids from the machine-learned potential energy landscape [68.25662704255433]
  We propose an approach for the quick and reliable screening of ionic conductors through the analysis of a universal interatomic potential.<n>Eight out of the ten highest-ranked materials are confirmed to be superionic at room temperature in first-principles calculations.<n>Our method achieves a speed-up factor of approximately 50 compared to molecular dynamics driven by a machine-learning potential, and is at least 3,000 times faster compared to first-principles molecular dynamics.
  arXiv  Detail & Related papers  (2024-11-11T09:01:36Z)
• Pre-trained Molecular Language Models with Random Functional Group Masking [54.900360309677794]
  We propose a SMILES-based underlineem Molecular underlineem Language underlineem Model, which randomly masking SMILES subsequences corresponding to specific molecular atoms. This technique aims to compel the model to better infer molecular structures and properties, thus enhancing its predictive capabilities.
  arXiv  Detail & Related papers  (2024-11-03T01:56:15Z)
• MolTRES: Improving Chemical Language Representation Learning for Molecular Property Prediction [14.353313239109337]
  MolTRES is a novel chemical language representation learning framework. It incorporates generator-discriminator training, allowing the model to learn from more challenging examples. Our model outperforms existing state-of-the-art models on popular molecular property prediction tasks.
  arXiv  Detail & Related papers  (2024-07-09T01:14:28Z)
• Balancing Molecular Information and Empirical Data in the Prediction of Physico-Chemical Properties [8.649679686652648]
  We propose a general method for combining molecular descriptors with representation learning.<n>The proposed hybrid model exploits chemical structure information using graph neural networks.<n>It automatically detects cases where structure-based predictions are unreliable, in which case it corrects them by representation-learning based predictions.
  arXiv  Detail & Related papers  (2024-06-12T10:51:00Z)
• Atomic and Subgraph-aware Bilateral Aggregation for Molecular Representation Learning [57.670845619155195]
  We introduce a new model for molecular representation learning called the Atomic and Subgraph-aware Bilateral Aggregation (ASBA) ASBA addresses the limitations of previous atom-wise and subgraph-wise models by incorporating both types of information. Our method offers a more comprehensive way to learn representations for molecular property prediction and has broad potential in drug and material discovery applications.
  arXiv  Detail & Related papers  (2023-05-22T00:56:00Z)
• Implicit Geometry and Interaction Embeddings Improve Few-Shot Molecular Property Prediction [53.06671763877109]
  We develop molecular embeddings that encode complex molecular characteristics to improve the performance of few-shot molecular property prediction. Our approach leverages large amounts of synthetic data, namely the results of molecular docking calculations. On multiple molecular property prediction benchmarks, training from the embedding space substantially improves Multi-Task, MAML, and Prototypical Network few-shot learning performance.
  arXiv  Detail & Related papers  (2023-02-04T01:32:40Z)
• Discovery of structure-property relations for molecules via hypothesis-driven active learning over the chemical space [0.0]
  We introduce a novel approach for the active learning over the chemical spaces based on hypothesis learning. We construct the hypotheses on the possible relationships between structures and functionalities of interest based on a small subset of data. This approach combines the elements from the symbolic regression methods such as SISSO and active learning into a single framework.
  arXiv  Detail & Related papers  (2023-01-06T14:22:43Z)
• Supervised Pretraining for Molecular Force Fields and Properties Prediction [16.86839767858162]
  We propose to pretrain neural networks on a dataset of 86 millions of molecules with atom charges and 3D geometries as inputs and molecular energies as labels. Experiments show that, compared to training from scratch, fine-tuning the pretrained model can significantly improve the performance for seven molecular property prediction tasks and two force field tasks.
  arXiv  Detail & Related papers  (2022-11-23T08:36:50Z)
• A Molecular Multimodal Foundation Model Associating Molecule Graphs with Natural Language [63.60376252491507]
  We propose a molecular multimodal foundation model which is pretrained from molecular graphs and their semantically related textual data. We believe that our model would have a broad impact on AI-empowered fields across disciplines such as biology, chemistry, materials, environment, and medicine.
  arXiv  Detail & Related papers  (2022-09-12T00:56:57Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.