IP Library › Granted Patent US 12,699,703
Granted Patent B2
US 12,699,703 · App. 19/484,186 · Granted Aug 4, 2026

Generative hierarchical searching for compounds

Inventors: Mengjiao Yang (Berkeley, CA); Simon Lutz Batzner (San Francisco, CA); Muratahan Aykol (San Jose, CA); Ruiqi Gao (San Francisco, CA); Alexander Lloyd Gaunt (Cambridgeshire, GB); Brendan Charles McMorrow (London, GB); Dale Eric Schuurmans (Edmonton, CA); Ekin Dogus Cubuk (Berkeley, CA); Igor Mordatch (Oakland, CA); Danilo Jimenez Rezende (London, GB)
Assignee: GDM Holding LLC
G06F16/24578G06F16/243G16C20/40G16C20/70
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,699,703
App. No.
19/484,186
Granted
Aug 4, 2026
Kind
B2
Abstract

A method performed by one or more computers. The method comprises receiving a natural language query specifying requirements for a compound; processing the natural language query using a language policy to generate a plurality of representations of candidate compounds that each satisfy at least a subset of the requirements specified in the natural language query. Each representation specifies at least a chemical formula of the corresponding candidate compound. The method further comprises, for each representation in a subset of the representations, using a generative machine learning model conditioned on the representation to generate one or more candidate chemical structures, each candidate chemical structure comprising a respective spatial location for each of the atoms of the corresponding candidate compound; and selecting a chemical structure and corresponding compound from the plurality of candidate chemical structures.

Claims (34)

1 . A method performed by one or more computers, the method comprising:

receiving a natural language query specifying requirements for a compound;

processing the natural language query using a language policy to generate a plurality of representations of candidate compounds that each satisfy at least a subset of the requirements specified in the natural language query, each representation specifying at least a chemical formula of the corresponding candidate compound;

for each representation in a subset of the representations, using a generative machine learning model conditioned on the representation to generate one or more candidate chemical structures, each candidate chemical structure comprising a respective spatial location for each of the atoms of the corresponding candidate compound; and

selecting a chemical structure and corresponding compound from the plurality of candidate chemical structures.

2 . The method of claim 1 , wherein the subset of the representations is selected using a first heuristic function that determines a respective likelihood that each of the representations satisfies the requirements specified in the natural language query.

3 . The method of claim 2 , wherein the first heuristic function comprises a pre-trained language model.

4 . The method of claim 3 , wherein the pre-trained language model is prompted to rank the representations according to the respective likelihood that each of the representations satisfies the requirements specified in the natural language query.

5 . The method of claim 1 , wherein the selected chemical structure and corresponding compound is selected from the plurality of candidate chemical structures using a second heuristic function, the selection being based on a prediction of one or more properties of each candidate chemical structure.

6 . The method of claim 5 , wherein the second heuristic function comprises a machine learning model that has been trained to process an input comprising a chemical structure to generate an output comprising a prediction of one or more properties of the chemical structure.

7 . The method of claim 6 , wherein the one or more properties of the chemical structure comprise one or more of: a formation energy; a bandgap energy; a conductivity; a magnetic property; an electrical property; a mechanical property; and a phase change property.

8 . The method of claim 7 , wherein the machine learning model is a graph neural network.

9 . The method of claim 1 , wherein selecting the chemical structure and corresponding compound from the plurality of candidate chemical structures comprises performing a tree search over the representations and candidate chemical structures.

10 . The method of claim 9 , wherein the tree search uses one of: a best-first search strategy; a breadth-first search strategy; and a depth-first search strategy.

11 . The method of claim 1 , wherein the language policy comprises a language model that processes the natural language query to generate the plurality of representations of candidate compounds.

12 . The method of claim 11 , wherein the language policy comprises a retrieval policy that processes the natural language query to retrieve data relating to the natural language query from one or more repositories, and wherein the language model is conditioned on the data retrieved from the one or more repositories.

13 . The method of claim 12 , wherein the data retrieved from the one or more repositories is added to the natural language query before the language model processes the natural language query.

14 . The method of claim 1 , wherein each representation identifies a point group or space group for the corresponding candidate chemical structure.

15 . The method of claim 1 , wherein each representation is a text representation or a multimodal representation that comprises text data.

16 . The method of claim 1 , wherein the natural language query comprises a plurality of known chemical formulas and the requirements specify that the compound is not one of the plurality of known chemical formulas.

17 . The method of claim 1 , wherein the requirements specify a chemical family for the compound.

18 . The method of claim 1 , wherein the requirements specify a stoichiometry or range of stoichiometries for the compound.

19 . One or more non-transitory computer storage media storing instructions that when executed by one or more computers cause the one or more computers to perform operations comprising:

receiving a natural language query specifying requirements for a compound;

processing the natural language query using a language policy to generate a plurality of representations of candidate compounds that each satisfy at least a subset of the requirements specified in the natural language query, each representation specifying at least a chemical formula of the corresponding candidate compound;

for each representation in a subset of the representations, using a generative machine learning model conditioned on the representation to generate one or more candidate chemical structures, each candidate chemical structure comprising a respective spatial location for each of the atoms of the corresponding candidate compound; and

selecting a chemical structure and corresponding compound from the plurality of candidate chemical structures.

20 . A system comprising:

one or more computers; and

one or more storage devices communicatively coupled to the one or more computers, wherein the one or more storage devices store instructions that, when executed by the one or more computers, cause the one or more computers to perform operations comprising:

receiving a natural language query specifying requirements for a compound;

processing the natural language query using a language policy to generate a plurality of representations of candidate compounds that each satisfy at least a subset of the requirements specified in the natural language query, each representation specifying at least a chemical formula of the corresponding candidate compound;

for each representation in a subset of the representations, using a generative machine learning model conditioned on the representation to generate one or more candidate chemical structures, each candidate chemical structure comprising a respective spatial location for each of the atoms of the corresponding candidate compound; and

selecting a chemical structure and corresponding compound from the plurality of candidate chemical structures.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 12, 2026
From: YANG, MENGJIAO; BATZNER, SIMON LUTZ; AYKOL, MURATAHAN; GAO, RUIQI; GAUNT, ALEXANDER LLOYD; MCMORROW, BRENDAN CHARLES; SCHUURMANS, DALE ERIC; CUBUK, EKIN DOGUS; MORDATCH, IGOR; REZENDE, DANILO JIMENEZ
To: GDM HOLDING LLC
Reel/Frame 074634/0308 →
Continuity (2)
Provisional Application 63650886 · May 22, 2024
Related Publication 20260147775A1 · May 28, 2026
References Cited (91)
US 5495604A · Harding · 1996 [cited by examiner]
US 11238101B1 · Mohajer · 2022 [cited by examiner]
US 12189590B1 · Fan · 2025 [cited by examiner]
US 20210057050A1 · Zavoronkovs · 2021 [cited by examiner]
US 20210191936A1 · Khorasani · 2021 [cited by examiner]
US 20210192131A1 · Levy · 2021 [cited by examiner]
US 20220101276A1 · Banatao · 2022 [cited by examiner]
US 20230098398A1 · Xu · 2023 [cited by examiner]
US 20230186025A1 · John · 2023 [cited by examiner]
US 20240029836A1 · Nie · 2024 [cited by examiner]
US 20240419835A1 · Karlberg · 2024 [cited by examiner]
US 20250166725A1 · Choi · 2025 [cited by examiner]
US 20250189496A1 · Healey · 2025 [cited by examiner]
US 20250225108A1 · Sharma · 2025 [cited by examiner]
US 20250225128A1 · Lazarev · 2025 [cited by examiner]
US 20250225377A1 · Gabellini · 2025 [cited by examiner]
US 20250232847A1 · Zavoronkovs · 2025 [cited by examiner]
US 20250284888A1 · Khullar · 2025 [cited by examiner]
US 20250321962A1 · Burger · 2025 [cited by examiner]
US 20250322916A1 · Chen · 2025 [cited by examiner]
US 20250328525A1 · Mishra · 2025 [cited by examiner]
US 20250328559A1 · Mey · 2025 [cited by examiner]
US 20250355897A1 · Sheetrit · 2025 [cited by examiner]
Ajay et al., “Compositional foundation models for hierarchical planning.” Advances in Neural Information Processing Systems 36, Dec. 2023, 22 pages. [cited by applicant]
Alayrac et al., “Flamingo: a Visual Language Model for Few-Shot Learning” CoRR, Submitted on Nov. 2022, arXiv:2204.14198v2, 54 pages. [cited by applicant]
Antunes et al., “Crystal structure generation with autoregressive large language modeling” CoRR, Submitted on Feb. 2024, arXiv:2307.04340v3, 55 pages. [cited by applicant]
Blochl et al., “Projector augmented-wave method” Physical review B, Dec. 1994, 50(24): 17953-17979. [cited by applicant]
Çiçek et al., “3d u-net: learning dense volumetric segmentation from sparse annotation” CoRR, Submitted on Jun. 2016, arXiv:1606.06650v1, 8 pages. [cited by applicant]
Court et al., “3-d inorganic crystal structure generation and property prediction via representation learning” Journal of Chemical Information and Modeling, Aug. 2020, 60(10): 4518-4535. [cited by applicant]
Cui et al., “A survey on multimodal large language models for autonomous driving” In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2024, pp. 958-979. [cited by applicant]
Davies et al., “Smact: Semiconducting materials by analogy and chemical theory.” Journal of Open Source Software 4.38, Jun. 2019, 1361: 3 pages. [cited by applicant]
Draxl et al., “The NOMAD laboratory: from data sharing to artificial intelligence.” Journal of Physics: Materials 2.3, May 2019, 036001: 9 pages. [cited by applicant]
Du et al., “Learning universal policies via text-guided video generation.” Advances in neural information processing systems 36, Dec. 2023, 17 pages. [cited by applicant]
Du et al., “Video language planning” CoRR, Submitted on Oct. 2023, arXiv:2310.10625v1, 17 pages. [cited by applicant]
Flam-Shepherd et al., “Language models can generate molecules, materials, and protein binding sites directly in three dimensions as XYZ, CIF, and PDB files” CoRR, Submitted on May 2023, arXiv:2305.05708v1, 14 pages. [cited by applicant]
Google Gemini Team et al., “Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context” CoRR, Submitted on Dec. 2024, arXiv:2403.05530v5, 154 pages. [cited by applicant]
Green et al., “The emergence of perovskite solar cells.” Nature photonics 8.7, Jul. 2014, 506-514. [cited by applicant]
Gruver et al., “Fine-tuned language models generate stable inorganic materials as text” CoRR, Submitted on Jul. 2025, arXiv:2402.04379v2, 20 pages. [cited by applicant]
Hellenbrandt, “The inorganic crystal structure database (ICSD)-present and future.” Crystallography Reviews 10.1, Jan. 2004, 17-22. [cited by applicant]
Ho et al., “Cascaded diffusion models for high fidelity image generation.” Journal of Machine Learning Research 23.47, 2022, 1-33. [cited by applicant]
Ho et al., “Imagen video: High definition video generation with diffusion models” CoRR, Submitted on Oct. 2022, arXiv:2210.02303v1, 18 pages. [cited by applicant]
Ho et al., “Video diffusion models.” Advances in neural information processing systems 35, Dec. 2022, 14 pages. [cited by applicant]
Hoffmann et al., “Data-Driven Approach to Encoding and Decoding 3-D Crystal Structures” CoRR, Submitted on Sep. 2019, arXiv:1909.00949v1, 21 pages. [cited by applicant]
Holmes et al., “Evaluating large language models on a highly-specialized topic, radiation oncology physics.” Frontiers in Oncology 13, Jul. 2023, 11 pages. [cited by applicant]
Ikebata et al., “Bayesian molecular design with a chemical language model.” Journal of computer-aided molecular design 31.4, Apr. 2017, 379-391. [cited by applicant]
Jain et al., “Commentary: The Materials Project: A materials genome approach to accelerating materials innovation.” APL materials 1.1, Jul. 2013, 011002: 1-11. [cited by applicant]
Kingma et al., “Adam: A Method for Stochastic Optimization” CoRR, Submitted on Jan. 2017, arXiv:1412.6980v9, 15 pages. [cited by applicant]
Kirklin et al., “The Open Quantum Materials Database (OQMD): assessing the accuracy of DFT formation energies.” npj Computational Materials 1.1, Dec. 2015, 1-15. [cited by applicant]
Kresse et al., “Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set.” Computational materials science 6.1, Jul. 1996, 15-50. [cited by applicant]
Kresse et al., “Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set.” Physical review B 54.16, Oct. 1996, 11169-11186. [cited by applicant]
Kresse et al., “From ultrasoft pseudopotentials to the projector augmented-wave method.” Physical review b 59.3, Jan. 1999, 1758-1774. [cited by applicant]
Lei et al., “Materials science in the era of large language models: a perspective” cond-mat, Submitted on Mar. 2024, arXiv:2403.06949v1, 33 pages. [cited by applicant]
Liang et al., “Code as Policies: Language Model Programs for Embodied Control” CoRR, Submitted on May 2023, arXiv:2209.07753v4, 16 pages. [cited by applicant]
Luu et al., “BioinspiredLLM: Conversational large language model for the mechanics of biological and bio-inspired materials.” Advanced Science 11.10, Mar. 2024, 2306724: 1-16. [cited by applicant]
Mathew et al., “Atomate: A high-level interface to generate, execute, and analyze computational materials science workflows.” Computational Materials Science 139, Nov. 2017, 22 pages. [cited by applicant]
Merchant et al., “Scaling deep learning for materials discovery.” Nature 624.7990, Dec. 2023, 80-85. [cited by applicant]
Mizushima et al., “LixCoO2 (0<x<−1): A new cathode material for batteries of high energy density.” Materials Research Bulletin 15.6, Jun. 1980, 783-789. [cited by applicant]
Momma et al., “VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data.” Applied Crystallography 44.6, Dec. 2011, 1272-1276. [cited by applicant]
Moret et al., “Leveraging molecular structure and bioactivity with chemical language models for de novo drug design.” Nature Communications 14.1, Jan. 2023, 114: 12 pages. [cited by applicant]
Noh et al., “Inverse design of solid-state materials via a continuous representation.” Matter 1.5, Nov. 2019, 1370-1384. [cited by applicant]
Ong et al., “Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis.” Computational Materials Science 68, Feb. 2013, 314-319. [cited by applicant]
Openai.com [online], “Video generation models as world simulators” Feb. 2024, retrieved on Nov. 17, 2025, retrieved from URL <https://openai.com/index/video-generation-models-as-world-simulators/>, 50 pages. [cited by applicant]
Peebles et al., “Scalable diffusion models with transformers.” Proceedings of the IEEE/CVF international conference on computer vision, 2023, 4195-4205. [cited by applicant]
Perdew et al., “Rationale for mixing exact exchange with density functional approximations.” The Journal of chemical physics 105.22, Dec. 1996, 9982-9985. [cited by applicant]
Ramesh et al., “Hierarchical Text-Conditional Image Generation with CLIP Latents” CoRR, Submitted on Apr. 2022, arXiv:2204.06125v1, 27 pages. [cited by applicant]
Ramesh et al., “Zero-shot text-to-image generation” Paper, Presented at International conference on machine learning, Jul. 2021, 11 pages. [cited by applicant]
Rombach et al., “High-resolution image synthesis with latent diffusion models.” Paper, Presented at Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, 10684-10695. [cited by applicant]
Ronneberger et al., “U-Net: Convolutional Networks for Biomedical Image Segmentation” CoRR, Submitted on May 2015, arXiv:1505.04597v1, 8 pages. [cited by applicant]
Saal et al., “Materials design and discovery with high-throughput density functional theory: the open quantum materials database (OQMD).” Jom 65.11, Nov. 2013, 1501-1509. [cited by applicant]
Saharia et al., “Photorealistic text-to-image diffusion models with deep language understanding.” Advances in neural information processing systems 35, Dec. 2022, 16 pages. [cited by applicant]
Savinov et al., “Step-unrolled Denoising Autoencoders for Text Generation” CoRR, Submitted on Apr. 2022, arXiv:2112.06749v3, 23 pages. [cited by applicant]
Sun et al., “Scieval: A multi-level large language model evaluation benchmark for scientific research.” Proceedings of the AAAI Conference on Artificial Intelligence. vol. 38. No. 17, Mar. 2024, 19053-19061. [cited by applicant]
Vahdat et al., “Score-based generative modeling in latent space.” Advances in neural information processing systems 34, Dec. 2021, 16 pages. [cited by applicant]
Valmeekam et al., “Large language models still can't plan (a benchmark for llms on planning and reasoning about change)” CoRR, Submitted on Apr. 2023, arXiv:2206.10498v3, 21 pages. [cited by applicant]
Valmeekam et al., “On the planning abilities of large language models—a critical investigation.” Advances in Neural Information Processing Systems 36, Dec. 2023, 13 pages. [cited by applicant]
Xie et al., “Translating natural language to planning goals with large-language models” CoRR, Submitted on Feb. 2023, arXiv:2302.05128v1, 15 pages. [cited by applicant]
Xie et al., “Crystal diffusion variational autoencoder for periodic material generation” CoRR, Submitted on Mar. 2022, arXiv:2110.06197v3, 20 pages. [cited by applicant]
Yang et al., “Learning interactive real-world simulators” CoRR, Submitted on Oct. 2023, arXiv:2310.06114v1, 26 pages. [cited by applicant]
Yang et al., “Scalable diffusion for materials generation” CoRR, Submitted on Jun. 2024, arXiv:2311.09235v2, 18 pages. [cited by applicant]
Zeni et al., “Mattergen: a generative model for inorganic materials design” CoRR, Submitted on Jan. 2024, arXiv:2312.03687v2, 56 pages. [cited by applicant]
Zhang et al., “Chemllm: A chemical large language model” CoRR, Submitted on Apr. 2042, arXiv:2402.06852v2, 26 pages. [cited by applicant]
Zhou et al., “Vision language models in autonomous driving and intelligent transportation systems” CoRR, Submitted on Oct. 2023, arXiv:2310.14414v1, 13 pages. [cited by applicant]
Bran et al., “Augmenting large language models with chemistry tools.” Nature Machine Intelligence 6.5 (2024): 525-535. [cited by applicant]
Edwards et al., “Text2mol: Cross-modal molecule retrieval with natural language queries.” Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Nov. 2021, 595-607. [cited by applicant]
Gong et al., “Text-Guided Molecule Generation with Diffusion Language Model” CoRR, Submitted on Feb. 2024, arXiv:2402.13040v1, 9 pages. [cited by applicant]
International Search Report and Written Opinion in International Appln. No. PCT/US2025/030389, mailed on Sep. 26, 2025, 20 pages. [cited by applicant]
Invitation to Pay Additional Fees in International Appln. No. PCT/US2025/030389, mailed on Aug. 5, 2025, 16 pages. [cited by applicant]
Li et al., “Empowering Molecule Discovery for Molecule-Caption Translation with Large Language Models: A ChatGPT Perspective” CoRR, Submitted on Apr. 2024, arXiv:2306.06615v2, 13 pages. [cited by applicant]
Su et al., “A Molecular Multimodal Foundation Model Associating Molecule Graphs with Natural Language” CoRR, Submitted on Sep. 2022, arXiv:2209.05481v1, 17 pages. [cited by applicant]
Ye, “De novo drug design as GPT language modeling: large chemistry models with supervised and reinforcement learning.” Journal of Computer-Aided Molecular Design 38.1, Dec. 2024, 20: 1-15. [cited by applicant]
Yirik et al., “MAYGEN: an open-source chemical structure generator for constitutional isomers based on the orderly generation principle.” Journal of cheminformatics 13.1, Jul. 2021, 48: 1-14. [cited by applicant]