PRISMat: Policy-Driven, Permutation-Invariant Autoregressive Material Generation

• URL: http://arxiv.org/abs/2605.16612v1
• Date: Fri, 15 May 2026 20:27:11 GMT
• Title: PRISMat: Policy-Driven, Permutation-Invariant Autoregressive Material Generation
• Abstract summary: Large Language Models (LLMs) are parameter-heavy and computationally expensive.<n>We present PRISMat, a cost-effective, permutation-invariant model.<n>We show that PRISMat, despite taking less time for inference, is able to outperform LLMs in generating crystal slabs conditioned on critical materials' surface properties.
• Score: 9.705199656826824
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Rapid identification of candidate materials with target properties has become a key task in materials science. Machine learning has emerged as an alternative to physics-based simulation, offering a faster and cheaper way to filter materials based on their stability and other target properties, reducing the number of candidates that reach the costly synthesis stage. Recently, Large Language Models (LLMs) have been applied to this role, but these models are parameter-heavy and computationally expensive both during training and at inference time, making them unsuitable for high-throughput tasks. This inefficiency stems from both the large over-parameterization of language models and the difficulty of framing material generation as a sequence learning problem. In this paper, we present PRISMat, a cost-effective, permutation-invariant model, which addresses these limitations. We show that PRISMat, despite taking less time for inference, is able to outperform LLMs in generating crystal slabs conditioned on critical materials' surface properties. In targeted material discovery, we achieve mean absolute errors of 0.188 eV/A$^2$ and 2.79 eV for cleavage energy and work function tasks, respectively, reducing the error of the next best model by 4$\times$.

Related papers

• Lightweight Multimodal LLM-Enabled Cost-Effective Defect Grading of Power Transmission Equipment [22.09101968097738]
  This paper introduces a novel defect grading framework based on multimodal large language model (MLLM)<n>By sending a secondary request to this model, a small number of chain of thought-based question-answer pairs (Q&As) are generated.<n>These high-quality interpretable Q&As are used to train Qwen3-VL-8B via Low-Rank Adaption-based supervised fine-tuning (SFT)
  arXiv  Detail & Related papers  (2026-03-25T16:59:13Z)
• PLaID++: A Preference Aligned Language Model for Targeted Inorganic Materials Design [0.1408690282236036]
  We introduce PLaID++, an LLM post-trained for stable and property-guided crystal generation.<n>We find that performance hinges on our crystallographic representation and reward formulation.
  arXiv  Detail & Related papers  (2025-09-08T18:57:57Z)
• A Materials Foundation Model via Hybrid Invariant-Equivariant Architectures [53.273077346444886]
  Machine learning interatomic potentials (MLIPs) can predict energy, force, and stress of materials.<n>A key design choice in MLIPs involves the trade-off between invariant and equivariant architectures.<n>HIENet is a hybrid invariant-equivariant materials interatomic potential model that integrates both invariant and equivariant message passing layers.
  arXiv  Detail & Related papers  (2025-02-25T18:01:05Z)
• Evaluating Language Models as Synthetic Data Generators [99.16334775127875]
  AgoraBench is a benchmark that provides standardized settings and metrics to evaluate LMs' data generation abilities.<n>Through synthesizing 1.26 million training instances using 6 LMs and training 99 student models, we uncover key insights about LMs' data generation capabilities.
  arXiv  Detail & Related papers  (2024-12-04T19:20:32Z)
• Is Tokenization Needed for Masked Particle Modelling? [8.79008927474707]
  Masked particle modeling (MPM) is a self-supervised learning scheme for constructing expressive representations of unordered sets. We improve MPM by addressing inefficiencies in the implementation and incorporating a more powerful decoder. We show that these new methods outperform the tokenized learning objective from the original MPM on a new test bed for foundation models for jets.
  arXiv  Detail & Related papers  (2024-09-19T09:12:29Z)
• Step-by-Step Unmasking for Parameter-Efficient Fine-tuning of Large Language Models [18.877891285367216]
  We introduce $textID3$, a novel selective PEFT method that calculates parameter importance continually.<n>We analytically show that $textID3$ reduces the number of gradient updates by a factor of two, enhancing computational efficiency.
  arXiv  Detail & Related papers  (2024-08-26T17:58:53Z)
• SciQu: Accelerating Materials Properties Prediction with Automated Literature Mining for Self-Driving Laboratories [0.7673339435080445]
  Assessing different material properties to predict specific attributes is a fundamental requirement for materials science-based applications. Our study addresses these challenges by leveraging machine learning to analyze material properties with greater precision and efficiency. By automating the data extraction process and using the extracted information to train machine learning models, our developed model, SciQu, optimize material properties.
  arXiv  Detail & Related papers  (2024-07-11T08:12:46Z)
• Advancing the Robustness of Large Language Models through Self-Denoised Smoothing [50.54276872204319]
  Large language models (LLMs) have achieved significant success, but their vulnerability to adversarial perturbations has raised considerable concerns. We propose to leverage the multitasking nature of LLMs to first denoise the noisy inputs and then to make predictions based on these denoised versions. Unlike previous denoised smoothing techniques in computer vision, which require training a separate model to enhance the robustness of LLMs, our method offers significantly better efficiency and flexibility.
  arXiv  Detail & Related papers  (2024-04-18T15:47:00Z)
• OmniJet-$α$: The first cross-task foundation model for particle physics [0.0]
  Foundation models are multi-dataset and multi-task machine learning methods that once pre-trained can be fine-tuned for a variety of downstream applications. We report significant progress on this challenge on several fronts. We demonstrate transfer learning between an unsupervised problem (jet generation) and a classic supervised task (jet tagging) with our new OmniJet-$alpha$ model.
  arXiv  Detail & Related papers  (2024-03-08T19:00:01Z)
• Leaving the Nest: Going Beyond Local Loss Functions for Predict-Then-Optimize [57.22851616806617]
  We show that our method achieves state-of-the-art results in four domains from the literature. Our approach outperforms the best existing method by nearly 200% when the localness assumption is broken.
  arXiv  Detail & Related papers  (2023-05-26T11:17:45Z)
• Model-Agnostic Multitask Fine-tuning for Few-shot Vision-Language Transfer Learning [59.38343286807997]
  We propose Model-Agnostic Multitask Fine-tuning (MAMF) for vision-language models on unseen tasks. Compared with model-agnostic meta-learning (MAML), MAMF discards the bi-level optimization and uses only first-order gradients. We show that MAMF consistently outperforms the classical fine-tuning method for few-shot transfer learning on five benchmark datasets.
  arXiv  Detail & Related papers  (2022-03-09T17:26:53Z)
• Polymers for Extreme Conditions Designed Using Syntax-Directed Variational Autoencoders [53.34780987686359]
  Machine learning tools are now commonly employed to virtually screen material candidates with desired properties. This approach is inefficient, and severely constrained by the candidates that human imagination can conceive. We utilize syntax-directed variational autoencoders (VAE) in tandem with Gaussian process regression (GPR) models to discover polymers expected to be robust under three extreme conditions.
  arXiv  Detail & Related papers  (2020-11-04T21:36:59Z)

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.