This paper provides a simulated laboratory for making use of Reinforcement Learning (RL) for chemical discovery. Since RL is fairly data intensive, training agents `on-the-fly' by taking actions in the real world is infeasible and possibly dangerous. Moreover, chemical processing and discovery involves challenges which are not commonly found in RL benchmarks and therefore offer a rich space to work in. We introduce a set of highly customizable and open-source RL environments, ChemGymRL, based on the standard Open AI Gym template. ChemGymRL supports a series of interconnected virtual chemical benches where RL agents can operate and train. The paper introduces and details each of these benches using well-known chemical reactions as illustrative examples, and trains a set of standard RL algorithms in each of these benches. Finally, discussion and comparison of the performances of several standard RL methods are provided in addition to a list of directions for future work as a vision for the further development and usage of ChemGymRL.

The properties of lithium metal are key parameters in the design of lithium ion and lithium metal batteries. They are difficult to probe experimentally due to the high reactivity and low melting point of lithium as well as the microscopic scales at which lithium exists in batteries where it is found to have enhanced strength, with implications for dendrite suppression strategies. Computationally, there is a lack of empirical potentials that are consistently quantitatively accurate across all properties and ab-initio calculations are too costly. In this work, we train Machine Learning Interaction Potentials (MLIPs) on Density Functional Theory (DFT) data to state-of-the-art accuracy in reproducing experimental and ab-initio results across a wide range of simulations at large length and time scales. We accurately predict thermodynamic properties, phonon spectra, temperature dependence of elastic constants and various surface properties inaccessible using DFT. We establish that there exists a Bell-Evans-Polanyi relation correlating the self-adsorption energy and the minimum surface diffusion barrier for high Miller index facets.

Timely feedback is an important part of teaching and learning. Here we describe how a readily available neural network transformer (machine-learning) model (BERT) can be used to give feedback on the structure of the response to an abstracting exercise where students are asked to summarise the contents of a published article after finding it from a publication database. The dataset contained 207 submissions from two consecutive years of the course, summarising a total of 21 different papers from the primary literature. The model was pre-trained using an available dataset (approx. 15,000 samples) and then fine-tuned on 80% of the submitted dataset. This fine tuning was seen to be important. The sentences in the student submissions are characterised into three classes - background, technique and observation - which allows a comparison of how each submission is structured. Comparing the structure of the students' abstract a large collection of those from the PubMed database shows that students in this exercise concentrate more on the background to the paper and less on the techniques and results than the abstracts to papers themselves. The results allowed feedback for each submitted assignment to be automatically generated.

We introduce a frustratingly simple, super efficient and surprisingly effective decoding method, which we call Frustratingly Simple Decoding (FSD), for neural text generation. The idea behind FSD is straightforward: we build an anti-LM based on previously generated text and use this anti-LM to penalize future generation of what has been generated. The anti-LM can be implemented as simple as an n-gram language model or a vectorized variant. In this way, FSD introduces no extra model parameters and negligible computational overhead (FSD can be as fast as greedy search). Despite the simplicity, FSD is surprisingly effective; Experiments show that FSD can outperform the canonical methods to date (i.e., nucleus sampling) as well as several strong baselines that were proposed recently.

Is there a unified framework for graph-based retrosynthesis prediction? Through analysis of full-, semi-, and non-template retrosynthesis methods, we discovered that they strive to strike an optimal balance between combinability and consistency: \textit{Should atoms be combined as motifs to simplify the molecular editing process, or should motifs be broken down into atoms to reduce the vocabulary and improve predictive consistency?} Recent works have studied several specific cases, while none of them explores different combinability-consistency trade-offs. Therefore, we propose MotifRetro, a dynamic motif editing framework for retrosynthesis prediction that can explore the entire trade-off space and unify graph-based models. MotifRetro comprises two components: RetroBPE, which controls the combinability-consistency trade-off, and a motif editing model, where we introduce a novel LG-EGAT module to dynamiclly add motifs to the molecule. We conduct extensive experiments on USPTO-50K to explore how the trade-off affects the model performance and finally achieve state-of-the-art performance.

Data-driven approaches for material discovery and design have been accelerated by emerging efforts in machine learning. However, general representations of crystals to explore the vast material search space remain limited. We introduce a material discovery framework that uses natural language embeddings derived from language models as representations of compositional and structural features. The discovery framework consists of a joint scheme that first recalls relevant candidates, and next ranks the candidates based on multiple target properties. The contextual knowledge encoded in language representations conveys information about material properties and structures, enabling both representational similarity analysis for recall, and multi-task learning to share information across related properties. By applying the framework to thermoelectrics, we demonstrate diversified recommendations of prototype structures and identify under-studied high-performance material spaces. The recommended materials are corroborated by first-principles calculations and experiments, revealing novel materials with potential high performance. Our framework provides a task-agnostic means for effective material recommendation and can be applied to various material systems.

Infusing deep learning with structural engineering has received widespread attention for both forward problems (structural simulation) and inverse problems (structural health monitoring). Based on Fourier Neural Operator, this study proposes VINO (Vehicle-bridge Interaction Neural Operator) to serve as the digital twin of bridge structures. VINO learns mappings between structural response fields and damage fields. In this study, VBI-FE dataset was established by running parametric finite element (FE) simulations considering a random distribution of structural initial damage field. Subsequently, VBI-EXP dataset was produced by conducting an experimental study under four damage scenarios. After VINO was pre-trained by VBI-FE and fine-tuned by VBI-EXP from the bridge at the healthy state, the model achieved the following two improvements. First, forward VINO can predict structural responses from damage field inputs more accurately than the FE model. Second, inverse VINO can determine, localize, and quantify damages in all scenarios, suggesting the practicality of data-driven approaches.

We present the latest release of PANNA 2.0 (Properties from Artificial Neural Network Architectures), a code for the generation of neural network interatomic potentials based on local atomic descriptors and multilayer perceptrons. Built on a new back end, this new release of PANNA features improved tools for customizing and monitoring network training, better GPU support including a fast descriptor calculator, new plugins for external codes and a new architecture for the inclusion of long-range electrostatic interactions through a variational charge equilibration scheme. We present an overview of the main features of the new code, and several benchmarks comparing the accuracy of PANNA models to the state of the art, on commonly used benchmarks as well as richer datasets.

Magnesium alloys are attractive options for temporary bio-implants because of their biocompatibility, controlled corrosion rate, and similarity to natural bone in terms of stiffness and density. Nevertheless, their low mechanical strength hinders their use as cardiovascular stents and bone substitutes. While it is possible to engineer alloys with the desired mechanical strength, optimizing the mechanical properties of biocompatible magnesium alloys using conventional experimental methods is time-consuming and expensive. Therefore, Artificial Intelligence (AI) can be leveraged to streamline the alloy design process and reduce the required time. In this study, a machine learning model was developed to predict the yield strength (YS) of biocompatible magnesium alloys with an $R^2$ accuracy of 91\%. The predictive model was then validated using the CALPHAD technique and thermodynamics calculations. Next, the predictive model was employed as the fitness function of a genetic algorithm to optimize the alloy composition for high-strength biocompatible magnesium implants. As a result, two alloys were proposed and synthesized, exhibiting YS values of 108 and 113 MPa, respectively. These values were substantially higher than those of conventional magnesium biocompatible alloys and closer to the YS and compressive strength of natural bone. Finally, the synthesized alloys were subjected to microstructure analysis and mechanical property testing to validate and evaluate the performance of the proposed AI-based alloy design approach for creating alloys with specific properties suitable for diverse applications.

Reaction diagram parsing is the task of extracting reaction schemes from a diagram in the chemistry literature. The reaction diagrams can be arbitrarily complex, thus robustly parsing them into structured data is an open challenge. In this paper, we present RxnScribe, a machine learning model for parsing reaction diagrams of varying styles. We formulate this structured prediction task with a sequence generation approach, which condenses the traditional pipeline into an end-to-end model. We train RxnScribe on a dataset of 1,378 diagrams and evaluate it with cross validation, achieving an 80.0% soft match F1 score, with significant improvements over previous models. Our code and data are publicly available at https://github.com/thomas0809/RxnScribe.