The Open Catalyst Challenge 2021: Competition Report

Abhishek Das, Muhammed Shuaibi, Aini Palizhati, Siddharth Goyal, Aditya Grover, Adeesh Kolluru, Janice Lan, Ammar Rizvi, Anuroop Sriram, Brandon Wood, Devi Parikh, Zachary Ulissi, C. Lawrence Zitnick, Guolin Ke, Shuxin Zheng, Yu Shi, Di He, Tie-Yan Liu, Chengxuan Ying, Jiacheng You, Yihan He, Rostislav Grigoriev, Ruslan Lukin, Adel Yarullin, Max Faleev
Proceedings of the NeurIPS 2021 Competitions and Demonstrations Track, PMLR 176:29-40, 2022.

Abstract

In this report, we describe the Open Catalyst Challenge held at NeurIPS 2021, focusing on using machine learning (ML) to accelerate the search for low-cost catalysts that can drive reactions converting renewable energy to storable forms. Specifically, the challenge required participants to develop ML approaches for relaxed energy prediction, i.e. given atomic positions for an adsorbate-catalyst system, the goal was to predict the energy of the system’s relaxed or lowest energy state. To perform well on this task, ML approaches need to approximate the quantum mechanical computations in Density Functional Theory (DFT). By modeling these accurately, the catalyst’s impact on the overall rate of a chemical reaction may be estimated; a key factor in filtering potential electrocatalyst materials. The challenge encouraged community-wide progress on this task and the winning approach improved direct relaxed energy prediction by  15% relative over the previous state-of-the-art.

Cite this Paper

BibTeX
@InProceedings{pmlr-v176-das22a, title = {The Open Catalyst Challenge 2021: Competition Report}, author = {Das, Abhishek and Shuaibi, Muhammed and Palizhati, Aini and Goyal, Siddharth and Grover, Aditya and Kolluru, Adeesh and Lan, Janice and Rizvi, Ammar and Sriram, Anuroop and Wood, Brandon and Parikh, Devi and Ulissi, Zachary and Zitnick, C. Lawrence and Ke, Guolin and Zheng, Shuxin and Shi, Yu and He, Di and Liu, Tie-Yan and Ying, Chengxuan and You, Jiacheng and He, Yihan and Grigoriev, Rostislav and Lukin, Ruslan and Yarullin, Adel and Faleev, Max}, booktitle = {Proceedings of the NeurIPS 2021 Competitions and Demonstrations Track}, pages = {29--40}, year = {2022}, editor = {Kiela, Douwe and Ciccone, Marco and Caputo, Barbara}, volume = {176}, series = {Proceedings of Machine Learning Research}, month = {06--14 Dec}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v176/das22a/das22a.pdf}, url = {https://proceedings.mlr.press/v176/das22a.html}, abstract = {In this report, we describe the Open Catalyst Challenge held at NeurIPS 2021, focusing on using machine learning (ML) to accelerate the search for low-cost catalysts that can drive reactions converting renewable energy to storable forms. Specifically, the challenge required participants to develop ML approaches for relaxed energy prediction, i.e. given atomic positions for an adsorbate-catalyst system, the goal was to predict the energy of the system’s relaxed or lowest energy state. To perform well on this task, ML approaches need to approximate the quantum mechanical computations in Density Functional Theory (DFT). By modeling these accurately, the catalyst’s impact on the overall rate of a chemical reaction may be estimated; a key factor in filtering potential electrocatalyst materials. The challenge encouraged community-wide progress on this task and the winning approach improved direct relaxed energy prediction by  15% relative over the previous state-of-the-art.} }
Endnote
%0 Conference Paper %T The Open Catalyst Challenge 2021: Competition Report %A Abhishek Das %A Muhammed Shuaibi %A Aini Palizhati %A Siddharth Goyal %A Aditya Grover %A Adeesh Kolluru %A Janice Lan %A Ammar Rizvi %A Anuroop Sriram %A Brandon Wood %A Devi Parikh %A Zachary Ulissi %A C. Lawrence Zitnick %A Guolin Ke %A Shuxin Zheng %A Yu Shi %A Di He %A Tie-Yan Liu %A Chengxuan Ying %A Jiacheng You %A Yihan He %A Rostislav Grigoriev %A Ruslan Lukin %A Adel Yarullin %A Max Faleev %B Proceedings of the NeurIPS 2021 Competitions and Demonstrations Track %C Proceedings of Machine Learning Research %D 2022 %E Douwe Kiela %E Marco Ciccone %E Barbara Caputo %F pmlr-v176-das22a %I PMLR %P 29--40 %U https://proceedings.mlr.press/v176/das22a.html %V 176 %X In this report, we describe the Open Catalyst Challenge held at NeurIPS 2021, focusing on using machine learning (ML) to accelerate the search for low-cost catalysts that can drive reactions converting renewable energy to storable forms. Specifically, the challenge required participants to develop ML approaches for relaxed energy prediction, i.e. given atomic positions for an adsorbate-catalyst system, the goal was to predict the energy of the system’s relaxed or lowest energy state. To perform well on this task, ML approaches need to approximate the quantum mechanical computations in Density Functional Theory (DFT). By modeling these accurately, the catalyst’s impact on the overall rate of a chemical reaction may be estimated; a key factor in filtering potential electrocatalyst materials. The challenge encouraged community-wide progress on this task and the winning approach improved direct relaxed energy prediction by  15% relative over the previous state-of-the-art.
APA
Das, A., Shuaibi, M., Palizhati, A., Goyal, S., Grover, A., Kolluru, A., Lan, J., Rizvi, A., Sriram, A., Wood, B., Parikh, D., Ulissi, Z., Zitnick, C.L., Ke, G., Zheng, S., Shi, Y., He, D., Liu, T., Ying, C., You, J., He, Y., Grigoriev, R., Lukin, R., Yarullin, A. & Faleev, M.. (2022). The Open Catalyst Challenge 2021: Competition Report. Proceedings of the NeurIPS 2021 Competitions and Demonstrations Track, in Proceedings of Machine Learning Research 176:29-40 Available from https://proceedings.mlr.press/v176/das22a.html.

Related Material