Search

The diffusional dynamics and vibrational spectroscopy of molecular hydrogen (H$_2$) in myoglobin (Mb) is characterized. Hydrogen has been implicated in a number of physiologically relevant processes, including cellular aging or inflammation. Here, the internal diffusion through the protein matrix was characterized and the vibrational spectroscopy was investigated using conventional empirical energy functions and improved models able to describe higher-order electrostatic moments of the ligand. H$_2$ can occupy the same internal defects as already found for Xe or CO (Xe1 to Xe4 and B-state). Furthermore, 4 additional sites were found, some of which had been discovered in earlier simulation studies. The vibrational spectra using the most refined energy function indicate that depending on the docking site the spectroscopy of H$_2$ differs. The maxima of the absorption spectra cover $\sim 20$ cm$^{-1}$ which are indicative of a pronounced effect of the surrounding protein matrix on the vibrational spectroscopy of the ligand. Electronic structure calculations show that H$_2$ forms a stable complex with the heme-iron (stabilized by $\sim -12$ kcal/mol) but splitting of H$_2$ is unlikely due to a high activation energy ($\sim 50$ kcal/mol).

Recent advances in eXplainable AI (XAI) for education have highlighted a critical challenge: ensuring that explanations for state-of-the-art AI models are understandable for non-technical users such as educators and students. In response, we introduce iLLuMinaTE, a zero-shot, chain-of-prompts LLM-XAI pipeline inspired by Miller's cognitive model of explanation. iLLuMinaTE is designed to deliver theory-driven, actionable feedback to students in online courses. iLLuMinaTE navigates three main stages - causal connection, explanation selection, and explanation presentation - with variations drawing from eight social science theories (e.g. Abnormal Conditions, Pearl's Model of Explanation, Necessity and Robustness Selection, Contrastive Explanation). We extensively evaluate 21,915 natural language explanations of iLLuMinaTE extracted from three LLMs (GPT-4o, Gemma2-9B, Llama3-70B), with three different underlying XAI methods (LIME, Counterfactuals, MC-LIME), across students from three diverse online courses. Our evaluation involves analyses of explanation alignment to the social science theory, understandability of the explanation, and a real-world user preference study with 114 university students containing a novel actionability simulation. We find that students prefer iLLuMinaTE explanations over traditional explainers 89.52% of the time. Our work provides a robust, ready-to-use framework for effectively communicating hybrid XAI-driven insights in education, with significant generalization potential for other human-centric fields.

While large language models (LLMs) are increasingly playing a pivotal role in education by providing instantaneous, adaptive responses, their potential to promote critical thinking remains understudied. In this paper, we fill such a gap and present an innovative educational chatbot designed to foster critical thinking through Socratic questioning. Unlike traditional intelligent tutoring systems, including educational chatbots, that tend to offer direct answers, the proposed Socratic tutor encourages students to explore various perspectives and engage in self-reflection by posing structured, thought-provoking questions. Our Socratic questioning is implemented by fine and prompt-tuning the open-source pretrained LLM with a specialized dataset that stimulates critical thinking and offers multiple viewpoints. In an effort to democratize access and to protect the students' privacy, the proposed tutor is based on small LLMs (Llama2 7B and 13B-parameter models) that are able to run locally on off-the-shelf hardware. We validate our approach in a battery of experiments consisting of interactions between a simulated student and the chatbot to evaluate its effectiveness in enhancing critical thinking skills. Results indicate that the Socratic tutor supports the development of reflection and critical thinking significantly better than standard chatbots. Our approach opens the door for improving educational outcomes by cultivating active learning and encouraging intellectual autonomy.

This work combines state-of-the-art machine learning techniques with highest-level electronic structure calculations and full-dimensional quantum tunneling calculations to obtain a quantitative characterization of tunneling splittings for system sizes that are currently out of reach using traditional approaches. For intramolecular hydrogen transfer in tropolone, the best computed splitting including perturbative corrections in the ring-polymer instanton calculations is 0.94 cm$^{-1}$ and compares with 0.974 cm$^{-1}$ from experiments. On the other hand, for intermolecular double hydrogen transfer in the (propiolic acid)-(formic acid) dimer, the computations yield 0.0147 cm$^{-1}$ which is larger by 40 % compared with experiment (0.0097 cm$^{-1}$) but still in much better agreement than previous attempts (0.63 cm$^{-1}$). The strategy pursued in the present work is applicable to yet larger systems and other properties of interest and provides a rational route for highest-accuracy energy functions for prediction and benchmarking electronic structure methods vis-a-vis experiments.

Intelligent Tutoring Systems (ITS) enhance personalized learning by predicting student answers to provide immediate and customized instruction. However, recent research has primarily focused on the correctness of the answer rather than the student's performance on specific answer choices, limiting insights into students' thought processes and potential misconceptions. To address this gap, we present MCQStudentBert, an answer forecasting model that leverages the capabilities of Large Language Models (LLMs) to integrate contextual understanding of students' answering history along with the text of the questions and answers. By predicting the specific answer choices students are likely to make, practitioners can easily extend the model to new answer choices or remove answer choices for the same multiple-choice question (MCQ) without retraining the model. In particular, we compare MLP, LSTM, BERT, and Mistral 7B architectures to generate embeddings from students' past interactions, which are then incorporated into a finetuned BERT's answer-forecasting mechanism. We apply our pipeline to a dataset of language learning MCQ, gathered from an ITS with over 10,000 students to explore the predictive accuracy of MCQStudentBert, which incorporates student interaction patterns, in comparison to correct answer prediction and traditional mastery-learning feature-based approaches. This work opens the door to more personalized content, modularization, and granular support., Comment: Accepted as a poster paper at EDM 2024: 17th International Conference on Educational Data Mining in Atlanta, USA

Clustering in education, particularly in large-scale online environments like MOOCs, is essential for understanding and adapting to diverse student needs. However, the effectiveness of clustering depends on its interpretability, which becomes challenging with high-dimensional data. Existing clustering approaches often neglect individual differences in feature importance and rely on a homogenized feature set. Addressing this gap, we introduce Interpret3C (Interpretable Conditional Computation Clustering), a novel clustering pipeline that incorporates interpretable neural networks (NNs) in an unsupervised learning context. This method leverages adaptive gating in NNs to select features for each student. Then, clustering is performed using the most relevant features per student, enhancing clusters' relevance and interpretability. We use Interpret3C to analyze the behavioral clusters considering individual feature importances in a MOOC with over 5,000 students. This research contributes to the field by offering a scalable, robust clustering methodology and an educational case study that respects individual student differences and improves interpretability for high-dimensional data., Comment: Accepted as a LBR paper at AIED 2024: The 25th International Conference on Artificial Intelligence in Education on July 8-12 in Recife, Brazil

There has been a growing interest in developing learner models to enhance learning and teaching experiences in educational environments. However, existing works have primarily focused on structured environments relying on meticulously crafted representations of tasks, thereby limiting the agent's ability to generalize skills across tasks. In this paper, we aim to enhance the generalization capabilities of agents in open-ended text-based learning environments by integrating Reinforcement Learning (RL) with Large Language Models (LLMs). We investigate three types of agents: (i) RL-based agents that utilize natural language for state and action representations to find the best interaction strategy, (ii) LLM-based agents that leverage the model's general knowledge and reasoning through prompting, and (iii) hybrid LLM-assisted RL agents that combine these two strategies to improve agents' performance and generalization. To support the development and evaluation of these agents, we introduce PharmaSimText, a novel benchmark derived from the PharmaSim virtual pharmacy environment designed for practicing diagnostic conversations. Our results show that RL-based agents excel in task completion but lack in asking quality diagnostic questions. In contrast, LLM-based agents perform better in asking diagnostic questions but fall short of completing the task. Finally, hybrid LLM-assisted RL agents enable us to overcome these limitations, highlighting the potential of combining RL and LLMs to develop high-performing agents for open-ended learning environments., Comment: Accepted as a full paper at EDM 2024: The 17th International Conference on Educational Data Mining, 14-17 of July 2024, Atlanta

Current course recommender systems primarily leverage learner-course interactions, course content, learner preferences, and supplementary course details like instructor, institution, ratings, and reviews, to make their recommendation. However, these systems often overlook a critical aspect: the evolving skill demand of the job market. This paper focuses on the perspective of academic researchers, working in collaboration with the industry, aiming to develop a course recommender system that incorporates job market skill demands. In light of the job market's rapid changes and the current state of research in course recommender systems, we outline essential properties for course recommender systems to address these demands effectively, including explainable, sequential, unsupervised, and aligned with the job market and user's goals. Our discussion extends to the challenges and research questions this objective entails, including unsupervised skill extraction from job listings, course descriptions, and resumes, as well as predicting recommendations that align with learner objectives and the job market and designing metrics to evaluate this alignment. Furthermore, we introduce an initial system that addresses some existing limitations of course recommender systems using large Language Models (LLMs) for skill extraction and Reinforcement Learning (RL) for alignment with the job market. We provide empirical results using open-source data to demonstrate its effectiveness., Comment: accepted at SIGIR 2024 as a perspective paper. Camera Ready will come soon

Recent advances in view synthesis and real-time rendering have achieved photorealistic quality at impressive rendering speeds. While Radiance Field-based methods achieve state-of-the-art quality in challenging scenarios such as in-the-wild captures and large-scale scenes, they often suffer from excessively high compute requirements linked to volumetric rendering. Gaussian Splatting-based methods, on the other hand, rely on rasterization and naturally achieve real-time rendering but suffer from brittle optimization heuristics that underperform on more challenging scenes. In this work, we present RadSplat, a lightweight method for robust real-time rendering of complex scenes. Our main contributions are threefold. First, we use radiance fields as a prior and supervision signal for optimizing point-based scene representations, leading to improved quality and more robust optimization. Next, we develop a novel pruning technique reducing the overall point count while maintaining high quality, leading to smaller and more compact scene representations with faster inference speeds. Finally, we propose a novel test-time filtering approach that further accelerates rendering and allows to scale to larger, house-sized scenes. We find that our method enables state-of-the-art synthesis of complex captures at 900+ FPS., Comment: Project page at https://m-niemeyer.github.io/radsplat/

Recent work exploring the capabilities of pre-trained large language models (LLMs) has demonstrated their ability to act as general pattern machines by completing complex token sequences representing a wide array of tasks, including time-series prediction and robot control. This paper investigates whether the pattern recognition and sequence modeling capabilities of LLMs can be extended to the domain of knowledge tracing, a critical component in the development of intelligent tutoring systems (ITSs) that tailor educational experiences by predicting learner performance over time. In an empirical evaluation across multiple real-world datasets, we compare two approaches to using LLMs for this task, zero-shot prompting and model fine-tuning, with existing, non-LLM approaches to knowledge tracing. While LLM-based approaches do not achieve state-of-the-art performance, fine-tuned LLMs surpass the performance of naive baseline models and perform on par with standard Bayesian Knowledge Tracing approaches across multiple metrics. These findings suggest that the pattern recognition capabilities of LLMs can be used to model complex learning trajectories, opening a novel avenue for applying LLMs to educational contexts. The paper concludes with a discussion of the implications of these findings for future research, suggesting that further refinements and a deeper understanding of LLMs' predictive mechanisms could lead to enhanced performance in knowledge tracing tasks., Comment: 12 pages, 4 figures

Uncertainty quantification (UQ) to detect samples with large expected errors (outliers) is applied to reactive molecular potential energy surfaces (PESs). Three methods - Ensembles, Deep Evidential Regression (DER), and Gaussian Mixture Models (GMM) - were applied to the H-transfer reaction between ${\it syn-}$Criegee and vinyl hydroxyperoxide. The results indicate that ensemble models provide the best results for detecting outliers, followed by GMM. For example, from a pool of 1000 structures with the largest uncertainty, the detection quality for outliers is $\sim 90$ \% and $\sim 50$ \%, respectively, if 25 or 1000 structures with large errors are sought. On the contrary, the limitations of the statistical assumptions of DER greatly impacted its prediction capabilities. Finally, a structure-based indicator was found to be correlated with large average error, which may help to rapidly classify new structures into those that provide an advantage for refining the neural network.

Interpretability for neural networks is a trade-off between three key requirements: 1) faithfulness of the explanation (i.e., how perfectly it explains the prediction), 2) understandability of the explanation by humans, and 3) model performance. Most existing methods compromise one or more of these requirements; e.g., post-hoc approaches provide limited faithfulness, automatically identified feature masks compromise understandability, and intrinsically interpretable methods such as decision trees limit model performance. These shortcomings are unacceptable for sensitive applications such as education and healthcare, which require trustworthy explanations, actionable interpretations, and accurate predictions. In this work, we present InterpretCC (interpretable conditional computation), a family of interpretable-by-design neural networks that guarantee human-centric interpretability, while maintaining comparable performance to state-of-the-art models by adaptively and sparsely activating features before prediction. We extend this idea into an interpretable, global mixture-of-experts (MoE) model that allows humans to specify topics of interest, discretely separates the feature space for each data point into topical subnetworks, and adaptively and sparsely activates these topical subnetworks for prediction. We apply variations of the InterpretCC architecture for text, time series and tabular data across several real-world benchmarks, demonstrating comparable performance with non-interpretable baselines, outperforming interpretable-by-design baselines, and showing higher actionability and usefulness according to a user study.

This survey article has grown out of the GAIED (pronounced "guide") workshop organized by the authors at the NeurIPS 2023 conference. We organized the GAIED workshop as part of a community-building effort to bring together researchers, educators, and practitioners to explore the potential of generative AI for enhancing education. This article aims to provide an overview of the workshop activities and highlight several future research directions in the area of GAIED.

The increasing availability of Massive Open Online Courses (MOOCs) has created a necessity for personalized course recommendation systems. These systems often combine neural networks with Knowledge Graphs (KGs) to achieve richer representations of learners and courses. While these enriched representations allow more accurate and personalized recommendations, explainability remains a significant challenge which is especially problematic for certain domains with significant impact such as education and online learning. Recently, a novel class of recommender systems that uses reinforcement learning and graph reasoning over KGs has been proposed to generate explainable recommendations in the form of paths over a KG. Despite their accuracy and interpretability on e-commerce datasets, these approaches have scarcely been applied to the educational domain and their use in practice has not been studied. In this work, we propose an explainable recommendation system for MOOCs that uses graph reasoning. To validate the practical implications of our approach, we conducted a user study examining user perceptions of our new explainable recommendations. We demonstrate the generalizability of our approach by conducting experiments on two educational datasets: COCO and Xuetang.

Large Language Models (LLMs) are increasingly utilized in educational tasks such as providing writing suggestions to students. Despite their potential, LLMs are known to harbor inherent biases which may negatively impact learners. Previous studies have investigated bias in models and data representations separately, neglecting the potential impact of LLM bias on human writing. In this paper, we investigate how bias transfers through an AI writing support pipeline. We conduct a large-scale user study with 231 students writing business case peer reviews in German. Students are divided into five groups with different levels of writing support: one classroom group with feature-based suggestions and four groups recruited from Prolific -- a control group with no assistance, two groups with suggestions from fine-tuned GPT-2 and GPT-3 models, and one group with suggestions from pre-trained GPT-3.5. Using GenBit gender bias analysis, Word Embedding Association Tests (WEAT), and Sentence Embedding Association Test (SEAT) we evaluate the gender bias at various stages of the pipeline: in model embeddings, in suggestions generated by the models, and in reviews written by students. Our results demonstrate that there is no significant difference in gender bias between the resulting peer reviews of groups with and without LLM suggestions. Our research is therefore optimistic about the use of AI writing support in the classroom, showcasing a context where bias in LLMs does not transfer to students' responses., Comment: Accepted as a full paper at EMNLP Findings 2023

Predicting multiple real-world tasks in a single model often requires a particularly diverse feature space. Multimodal (MM) models aim to extract the synergistic predictive potential of multiple data types to create a shared feature space with aligned semantic meaning across inputs of drastically varying sizes (i.e. images, text, sound). Most current MM architectures fuse these representations in parallel, which not only limits their interpretability but also creates a dependency on modality availability. We present MultiModN, a multimodal, modular network that fuses latent representations in a sequence of any number, combination, or type of modality while providing granular real-time predictive feedback on any number or combination of predictive tasks. MultiModN's composable pipeline is interpretable-by-design, as well as innately multi-task and robust to the fundamental issue of biased missingness. We perform four experiments on several benchmark MM datasets across 10 real-world tasks (predicting medical diagnoses, academic performance, and weather), and show that MultiModN's sequential MM fusion does not compromise performance compared with a baseline of parallel fusion. By simulating the challenging bias of missing not-at-random (MNAR), this work shows that, contrary to MultiModN, parallel fusion baselines erroneously learn MNAR and suffer catastrophic failure when faced with different patterns of MNAR at inference. To the best of our knowledge, this is the first inherently MNAR-resistant approach to MM modeling. In conclusion, MultiModN provides granular insights, robustness, and flexibility without compromising performance., Comment: Accepted as a full paper at NeurIPS 2023 in New Orleans, USA

The effect of noise in the input data for learning potential energy surfaces (PESs) based on neural networks for chemical applications is assessed. Noise in energies and forces can result from aleatoric and epistemic errors in the quantum chemical reference calculations. Statistical (aleatoric) noise arises for example due to the need to set convergence thresholds in the self consistent field (SCF) iterations whereas systematic (epistemic) noise is due to, {\it inter alia}, particular choices of basis sets in the calculations. The two molecules considered here as proxies are H$_{2}$CO and HONO which are examples for single- and multi-reference problems, respectively, for geometries around the minimum energy structure. For H$_2$CO it is found that adding noise to energies with magnitudes representative of single-point calculations does not deteriorate the quality of the final PESs whereas increasing the noise level commensurate with electronic structure calculations for more complicated, e.g. metal-containing, systems is expected to have a more notable effect. However, the effect of noise on the forces is more noticeable. On the other hand, for HONO which requires a multi-reference treatment, a clear correlation between model quality and the degree of multi-reference character as measured by the $T_1$ amplitude is found. It is concluded that for chemically "simple" cases the effect of aleatoric and epistemic noise is manageable without evident deterioration of the trained model - although the quality of the forces is important. However, considerably more care needs to be exercised for situations in which multi-reference effects are present.

Photoproduction of $\pi^0\pi^{\pm}$-pairs from quasifree nucleons bound in the deuteron has been investigated to study the helicity dependence of this reaction. Measurements with a liquid deuterium target were used to extract the unpolarized cross sections for reactions on protons and neutrons. A deuterated, longitudinally polarized solid-butanol target, together with a circularly polarized photon beam, determined the double polarization observable $E$. From these results the spin-dependent cross sections $\sigma_{1/2}$ and $\sigma_{3/2}$, corresponding to the anti-parallel and parallel spin configurations of the beam photon and target nucleon, have been derived. The measurements were performed at the Mainz MAMI accelerator with tagged, circularly-polarized photon beams produced via bremsstrahlung from longitudinally polarized electron beams. The reaction products were detected with an almost $4\pi$ solid-angle covering calorimeter composed of the Crystal Ball and TAPS detectors, supplemented by plastic scintillation detectors for charged particle identification. The results are sensitive to sequential decays of nucleon resonances via intermediate states and also to the decay of nucleon resonances by emission of charged $\rho$ mesons, and are compared to recent model results., Comment: 9 pages, 6 figures

Explainable Artificial Intelligence (XAI) plays a crucial role in enabling human understanding and trust in deep learning systems. As models get larger, more ubiquitous, and pervasive in aspects of daily life, explainability is necessary to minimize adverse effects of model mistakes. Unfortunately, current approaches in human-centric XAI (e.g. predictive tasks in healthcare, education, or personalized ads) tend to rely on a single post-hoc explainer, whereas recent work has identified systematic disagreement between post-hoc explainers when applied to the same instances of underlying black-box models. In this paper, we therefore present a call for action to address the limitations of current state-of-the-art explainers. We propose a shift from post-hoc explainability to designing interpretable neural network architectures. We identify five needs of human-centric XAI (real-time, accurate, actionable, human-interpretable, and consistent) and propose two schemes for interpretable-by-design neural network workflows (adaptive routing with InterpretCC and temporal diagnostics with I2MD). We postulate that the future of human-centric XAI is neither in explaining black-boxes nor in reverting to traditional, interpretable models, but in neural networks that are intrinsically interpretable., Comment: Viewpoint paper, under review at JAIR

Natural Language Feedback (NLF) is an increasingly popular mechanism for aligning Large Language Models (LLMs) to human preferences. Despite the diversity of the information it can convey, NLF methods are often hand-designed and arbitrary, with little systematic grounding. At the same time, research in learning sciences has long established several effective feedback models. In this opinion piece, we compile ideas from pedagogy to introduce FELT, a feedback framework for LLMs that outlines various characteristics of the feedback space, and a feedback content taxonomy based on these variables, providing a general mapping of the feedback space. In addition to streamlining NLF designs, FELT also brings out new, unexplored directions for research in NLF. We make our taxonomy available to the community, providing guides and examples for mapping our categorizations to future research., Comment: EMNLP 2024; 9 pages, 3 figures

Revision behavior in adaptive writing support systems is an important and relatively new area of research that can improve the design and effectiveness of these tools, and promote students' self-regulated learning (SRL). Understanding how these tools are used is key to improving them to better support learners in their writing and learning processes. In this paper, we present a novel pipeline with insights into the revision behavior of students at scale. We leverage a data set of two groups using an adaptive writing support tool in an educational setting. With our novel pipeline, we show that the tool was effective in promoting revision among the learners. Depending on the writing feedback, we were able to analyze different strategies of learners when revising their texts, we found that users of the exemplary case improved over time and that females tend to be more efficient. Our research contributes a pipeline for measuring SRL behaviors at scale in writing tasks (i.e., engagement or revision behavior) and informs the design of future adaptive writing support systems for education, with the goal of enhancing their effectiveness in supporting student writing. The source code is available at https://github.com/lucamouchel/Understanding-Revision-Behavior., Comment: 8 pages, Conference Paper

Flipped Classrooms (FC) are a promising teaching strategy, where students engage with the learning material before attending face-to-face sessions. While pre-class activities are critical for course success, many students struggle to engage effectively in them due to inadequate of self-regulated learning (SRL) skills. Thus, tools enabling teachers to monitor students' SRL and provide personalized guidance have the potential to improve learning outcomes. However, existing dashboards mostly focus on aggregated information, disregarding recent work leveraging machine learning (ML) approaches that have identified comprehensive, multi-dimensional SRL behaviors. Unfortunately, the complexity of such findings makes them difficult to communicate and act on. In this paper, we follow a teacher-centered approach to study how to make thorough findings accessible to teachers. We design and implement FlippED, a dashboard for monitoring students' SRL behavior. We evaluate the usability and actionability of the tool in semi-structured interviews with ten university teachers. We find that communicating ML-based profiles spark a range of potential interventions for students and course modifications., Comment: Accepted as a poster paper at AIED 2023: The 24th International Conference on Artificial Intelligence in Education, 3-7 of July 2023, Tokyo

Full dimensional potential energy surfaces (PESs) based on machine learning (ML) techniques provide means for accurate and efficient molecular simulations in the gas- and condensed-phase for various experimental observables ranging from spectroscopy to reaction dynamics. Here, the MLpot extension with PhysNet as the ML-based model for a PES is introduced into the newly developed pyCHARMM API. To illustrate conceiving, validating, refining and using a typical workflow, para-chloro-phenol is considered as an example. The main focus is on how to approach a concrete problem from a practical perspective and applications to spectroscopic observables and the free energy for the -OH torsion in solution are discussed in detail. For the computed IR spectra in the fingerprint region the computations for para-chloro-phenol in water are in good qualitative agreement with experiment carried out in CCl$_4$. Also, relative intensities are largely consistent with experimental findings. The barrier for rotation of the -OH group increases from $\sim 3.5$ kcal/mol in the gas phase to $\sim 4.1$ kcal/mol from simulations in water due to favourable H-bonding interactions of the -OH group with surrounding water molecules., Comment: 38 pages, 11 figures

Recent advances in experimental methodology enabled studies of the quantum-state and conformational dependence of chemical reactions under precisely controlled conditions in the gas phase. Here, we generated samples of selected gauche and s-trans 2,3-dibromobutadiene (DBB) by electrostatic deflection in a molecular beam and studied their reaction with Coulomb crystals of laser-cooled $\mathrm{Ca^{+}}$ ions in an ion trap. The rate coefficients for the total reaction were found to strongly depend on both the conformation of DBB and the electronic state of $\mathrm{Ca^{+}}$. In the $\mathrm{(4p)~^{2}P_{1/2}}$ and $\mathrm{(3d)~^{2}D_{3/2}}$ excited states of $\mathrm{Ca^{+}}$, the reaction is capture-limited and faster for the gauche conformer due to long-range ion-dipole interactions. In the $\mathrm{(4s)~^{2}S_{1/2}}$ ground state of $\mathrm{Ca^{+}}$, the reaction rate for s-trans DBB still conforms with the capture limit, while that for gauche DBB is strongly suppressed. The experimental observations were analysed with the help of adiabatic capture theory, ab-initio calculations and reactive molecular dynamics simulations on a machine-learned full-dimensional potential energy surface of the system. The theory yields near-quantitative agreement for s-trans-DBB, but overestimates the reactivity of the gauche-conformer compared to the experiment. The present study points to the important role of molecular geometry even in strongly reactive exothermic systems and illustrates striking differences in the reactivity of individual conformers in gas-phase ion-molecule reactions.

The rise of machine learning has greatly influenced the field of computational chemistry, and that of atomistic molecular dynamics simulations in particular. One of its most exciting prospects is the development of accurate, full-dimensional potential energy surfaces (PESs) for molecules and clusters, which, however, often require thousands to tens of thousands of ab initio data points restricting the community to medium sized molecules and/or lower levels of theory (e.g. DFT). Transfer learning, which improves a global PES from a lower to a higher level of theory, offers a data efficient alternative requiring only a fraction of the high level data (on the order of 100 are found to be sufficient for malonaldehyde). The present work demonstrates that even with Hartree-Fock theory and a double-zeta basis set as the lower level model, transfer learning yields CCSD(T)-level quality for H-transfer barrier energies, harmonic frequencies and H-transfer tunneling splittings. Most importantly, finite-temperature molecular dynamics simulations on the sub-microsecond time scale in the gas phase are possible and the infrared spectra determined from the transfer learned PESs are in good agreement with experiment. It is concluded that routine, long-time atomistic simulations on PESs fulfilling CCSD(T)-standards become possible.

Reasoning in a complex and ambiguous environment is a key goal for Reinforcement Learning (RL) agents. While some sophisticated RL agents can successfully solve difficult tasks, they require a large amount of training data and often struggle to generalize to new unseen environments and new tasks. On the other hand, Large Scale Language Models (LSLMs) have exhibited strong reasoning ability and the ability to to adapt to new tasks through in-context learning. However, LSLMs do not inherently have the ability to interrogate or intervene on the environment. In this work, we investigate how to combine these complementary abilities in a single system consisting of three parts: a Planner, an Actor, and a Reporter. The Planner is a pre-trained language model that can issue commands to a simple embodied agent (the Actor), while the Reporter communicates with the Planner to inform its next command. We present a set of tasks that require reasoning, test this system's ability to generalize zero-shot and investigate failure cases, and demonstrate how components of this system can be trained with reinforcement-learning to improve performance., Comment: Presented at NeurIPS 2022 Language and Reinforcement Learning Workshop (best paper) and NeurIPS 2022 Foundation Models for Decision Making Workshop. 4 pages main; 14 pages total (including references and appendix); 3 figures

Algorithms deployed in education can shape the learning experience and success of a student. It is therefore important to understand whether and how such algorithms might create inequalities or amplify existing biases. In this paper, we analyze the fairness of models which use behavioral data to identify at-risk students and suggest two novel pre-processing approaches for bias mitigation. Based on the concept of intersectionality, the first approach involves intelligent oversampling on combinations of demographic attributes. The second approach does not require any knowledge of demographic attributes and is based on the assumption that such attributes are a (noisy) proxy for student behavior. We hence propose to directly oversample different types of behaviors identified in a cluster analysis. We evaluate our approaches on data from (i) an open-ended learning environment and (ii) a flipped classroom course. Our results show that both approaches can mitigate model bias. Directly oversampling on behavior is a valuable alternative, when demographic metadata is not available. Source code and extended results are provided in https://github.com/epfl-ml4ed/behavioral-oversampling}{https://github.com/epfl-ml4ed/behavioral-oversampling ., Comment: Accepted as a full paper at LAK 2023: The 13th International Learning Analytics and Knowledge Conference, 13-17 of March 2023, Arlington

Deep learning models for learning analytics have become increasingly popular over the last few years; however, these approaches are still not widely adopted in real-world settings, likely due to a lack of trust and transparency. In this paper, we tackle this issue by implementing explainable AI methods for black-box neural networks. This work focuses on the context of online and blended learning and the use case of student success prediction models. We use a pairwise study design, enabling us to investigate controlled differences between pairs of courses. Our analyses cover five course pairs that differ in one educationally relevant aspect and two popular instance-based explainable AI methods (LIME and SHAP). We quantitatively compare the distances between the explanations across courses and methods. We then validate the explanations of LIME and SHAP with 26 semi-structured interviews of university-level educators regarding which features they believe contribute most to student success, which explanations they trust most, and how they could transform these insights into actionable course design decisions. Our results show that quantitatively, explainers significantly disagree with each other about what is important, and qualitatively, experts themselves do not agree on which explanations are most trustworthy. All code, extended results, and the interview protocol are provided at https://github.com/epfl-ml4ed/trusting-explainers., Comment: Accepted as a full paper (Best Paper nominee) at LAK 2023: The 13th International Learning Analytics and Knowledge Conference, March 13-17, 2023, Arlington, Texas, USA

Student success models might be prone to develop weak spots, i.e., examples hard to accurately classify due to insufficient representation during model creation. This weakness is one of the main factors undermining users' trust, since model predictions could for instance lead an instructor to not intervene on a student in need. In this paper, we unveil the need of detecting and characterizing unknown unknowns in student success prediction in order to better understand when models may fail. Unknown unknowns include the students for which the model is highly confident in its predictions, but is actually wrong. Therefore, we cannot solely rely on the model's confidence when evaluating the predictions quality. We first introduce a framework for the identification and characterization of unknown unknowns. We then assess its informativeness on log data collected from flipped courses and online courses using quantitative analyses and interviews with instructors. Our results show that unknown unknowns are a critical issue in this domain and that our framework can be applied to support their detection. The source code is available at https://github.com/epfl-ml4ed/unknown-unknowns., Comment: Accepted as a full paper at the International Conference on Learning Analytics & Knowledge (LAK23)

Time series is the most prevalent form of input data for educational prediction tasks. The vast majority of research using time series data focuses on hand-crafted features, designed by experts for predictive performance and interpretability. However, extracting these features is labor-intensive for humans and computers. In this paper, we propose an approach that utilizes irregular multivariate time series modeling with graph neural networks to achieve comparable or better accuracy with raw time series clickstreams in comparison to hand-crafted features. Furthermore, we extend concept activation vectors for interpretability in raw time series models. We analyze these advances in the education domain, addressing the task of early student performance prediction for downstream targeted interventions and instructional support. Our experimental analysis on 23 MOOCs with millions of combined interactions over six behavioral dimensions show that models designed with our approach can (i) beat state-of-the-art educational time series baselines with no feature extraction and (ii) provide interpretable insights for personalized interventions. Source code: https://github.com/epfl-ml4ed/ripple/., Comment: Accepted as a full paper at AAAI 2023: 37th AAAI Conference on Artificial Intelligence (EAAI: AI for Education Special Track), 7-14 of February 2023, Washington DC, USA