Christopher J. MacLellan

Darryl Hannan, Steven C. Nesbit, Ximing Wen et al.

Point-of-Care Ultrasound (POCUS) refers to clinician-performed and interpreted ultrasonography at the patient's bedside. Interpreting these images requires a high level of expertise, which may not be available during emergencies. In this paper, we support POCUS by developing classifiers that can aid medical professionals by diagnosing whether or not a patient has pneumothorax. We decomposed the task into multiple steps, using YOLOv4 to extract relevant regions of the video and a 3D sparse coding model to represent video features. Given the difficulty in acquiring positive training videos, we trained a small-data classifier with a maximum of 15 positive and 32 negative examples. To counteract this limitation, we leveraged subject matter expert (SME) knowledge to limit the hypothesis space, thus reducing the cost of data collection. We present results using two lung ultrasound datasets and demonstrate that our model is capable of achieving performance on par with SMEs in pneumothorax identification. We then developed an iOS application that runs our full system in less than 4 seconds on an iPad Pro, and less than 8 seconds on an iPhone 13 Pro, labeling key regions in the lung sonogram to provide interpretable diagnoses.

Christopher James MacLellan, Kimberly Stowers, Lisa Brady · gatech

Evaluating different training interventions to determine which produce the best learning outcomes is one of the main challenges faced by instructional designers. Typically, these designers use A/B experiments to evaluate each intervention; however, it is costly and time consuming to run such studies. To address this issue, we explore how computational models of learning might support designers in reasoning causally about alternative interventions within a fractions tutor. We present an approach for automatically tuning models to specific individuals and show that personalized models make better predictions of students' behavior than generic ones. Next, we conduct simulations to generate counterfactual predictions of performance and learning for two students (high and low performing) in different versions of the fractions tutor. Our approach makes predictions that align with previous human findings, as well as testable predictions that might be evaluated with future human experiments.

Jieyu Zhou, Aryan Roy, Sneh Gupta et al.

Existing AI agents typically execute multi-step tasks autonomously and only allow user confirmation at the end. During execution, users have little control, making the confirm-at-end approach brittle: a single error can cascade and force a complete restart. Confirming every step avoids such failures, but imposes tedious overhead. Balancing excessive interruptions against costly rollbacks remains an open challenge. We address this problem by modeling confirmation as a minimum time scheduling problem. We conducted a formative study with eight participants, which revealed a recurring Confirmation-Diagnosis-Correction-Redo (CDCR) pattern in how users monitor errors. Based on this pattern, we developed a decision-theoretic model to determine time-efficient confirmation point placement. We then evaluated our approach using a within-subjects study where 48 participants monitored AI agents and repaired their mistakes while executing tasks. Results show that 81 percent of participants preferred our intermediate confirmation approach over the confirm-at-end approach used by existing systems, and task completion time was reduced by 13.54 percent.

Lane Lawley, Christopher J. MacLellan

Machine learning often requires millions of examples to produce static, black-box models. In contrast, interactive task learning (ITL) emphasizes incremental knowledge acquisition from limited instruction provided by humans in modalities such as natural language. However, ITL systems often suffer from brittle, error-prone language parsing, which limits their usability. Large language models (LLMs) are resistant to brittleness but are not interpretable and cannot learn incrementally. We present VAL, an ITL system with a new philosophy for LLM/symbolic integration. By using LLMs only for specific tasks--such as predicate and argument selection--within an algorithmic framework, VAL reaps the benefits of LLMs to support interactive learning of hierarchical task knowledge from natural language. Acquired knowledge is human interpretable and generalizes to support execution of novel tasks without additional training. We studied users' interactions with VAL in a video game setting, finding that most users could successfully teach VAL using language they felt was natural.

Ximing Wen, Rosina O. Weber, Anik Sen et al.

Point-of-Care Ultrasound (POCUS) is the practice of clinicians conducting and interpreting ultrasound scans right at the patient's bedside. However, the expertise needed to interpret these images is considerable and may not always be present in emergency situations. This reality makes algorithms such as machine learning classifiers extremely valuable to augment human decisions. POCUS devices are becoming available at a reasonable cost in the size of a mobile phone. The challenge of turning POCUS devices into life-saving tools is that interpretation of ultrasound images requires specialist training and experience. Unfortunately, the difficulty to obtain positive training images represents an important obstacle to building efficient and accurate classifiers. Hence, the problem we try to investigate is how to explore strategies to increase accuracy of classifiers trained with scarce data. We hypothesize that training with a few data instances may not suffice for classifiers to generalize causing them to overfit. Our approach uses an Explainable AI-Augmented approach to help the algorithm learn more from less and potentially help the classifier better generalize.

Christopher J. MacLellan, Erik Harpstead, Vincent Aleven et al. · gatech

The literature on concept formation has demonstrated that humans are capable of learning concepts incrementally, with a variety of attribute types, and in both supervised and unsupervised settings. Many models of concept formation focus on a subset of these characteristics, but none account for all of them. In this paper, we present TRESTLE, an incremental account of probabilistic concept formation in structured domains that unifies prior concept learning models. TRESTLE works by creating a hierarchical categorization tree that can be used to predict missing attribute values and cluster sets of examples into conceptually meaningful groups. It updates its knowledge by partially matching novel structures and sorting them into its categorization tree. Finally, the system supports mixed-data representations, including nominal, numeric, relational, and component attributes. We evaluate TRESTLE's performance on a supervised learning task and an unsupervised clustering task. For both tasks, we compare it to a nonincremental model and to human participants. We find that this new categorization model is competitive with the nonincremental approach and more closely approximates human behavior on both tasks. These results serve as an initial demonstration of TRESTLE's capabilities and show that, by taking key characteristics of human learning into account, it can better model behavior than approaches that ignore them.

Zekun Wang, Anant Gupta, Christopher J. MacLellan

Continual learning aims to acquire tasks sequentially without catastrophic forgetting, yet standard strategies face a core tradeoff: regularization-based methods (e.g., EWC) can overconstrain updates when task optima are weakly overlapping, while replay-based methods can retain performance but drift due to imperfect replay. We study a hybrid perspective: \emph{trust region continual learning} that combines generative replay with a Fisher-metric trust region constraint. We show that, under local approximations, the resulting update admits a MAML-style interpretation with a single implicit inner step: replay supplies an old-task gradient signal (query-like), while the Fisher-weighted penalty provides an efficient offline curvature shaping (support-like). This yields an emergent meta-learning property in continual learning: the model becomes an initialization that rapidly \emph{re-converges} to prior task optima after each task transition, without explicitly optimizing a bilevel objective. Empirically, on task-incremental diffusion image generation and continual diffusion-policy control, trust region continual learning achieves the best final performance and retention, and consistently recovers early-task performance faster than EWC, replay, and continual meta-learning baselines.

Tommaso Calo, Christopher J. MacLellan

Intelligent Tutoring Systems (ITSs) have shown great potential in delivering personalized and adaptive education, but their widespread adoption has been hindered by the need for specialized programming and design skills. Existing approaches overcome the programming limitations with no-code authoring through drag and drop, however they assume that educators possess the necessary skills to design effective and engaging tutor interfaces. To address this assumption we introduce generative AI capabilities to assist educators in creating tutor interfaces that meet their needs while adhering to design principles. Our approach leverages Large Language Models (LLMs) and prompt engineering to generate tutor layout and contents based on high-level requirements provided by educators as inputs. However, to allow them to actively participate in the design process, rather than relying entirely on AI-generated solutions, we allow generation both at the entire interface level and at the individual component level. The former provides educators with a complete interface that can be refined using direct manipulation, while the latter offers the ability to create specific elements to be added to the tutor interface. A small-scale comparison shows the potential of our approach to enhance the efficiency of tutor interface design. Moving forward, we raise critical questions for assisting educators with generative AI capabilities to create personalized, effective, and engaging tutors, ultimately enhancing their adoption.

Adit Gupta, Jennifer Reddig, Tommaso Calo et al.

Researchers have made notable progress in applying Large Language Models (LLMs) to solve math problems, as demonstrated through efforts like GSM8k, ProofNet, AlphaGeometry, and MathOdyssey. This progress has sparked interest in their potential use for tutoring students in mathematics. However, the reliability of LLMs in tutoring contexts -- where correctness and instructional quality are crucial -- remains underexplored. Moreover, LLM problem-solving capabilities may not necessarily translate into effective tutoring support for students. In this work, we present two novel approaches to evaluate the correctness and quality of LLMs in math tutoring contexts. The first approach uses an intelligent tutoring system for college algebra as a testbed to assess LLM problem-solving capabilities. We generate benchmark problems using the tutor, prompt a diverse set of LLMs to solve them, and compare the solutions to those generated by the tutor. The second approach evaluates LLM as tutors rather than problem solvers. We employ human evaluators, who act as students seeking tutoring support from each LLM. We then assess the quality and correctness of the support provided by the LLMs via a qualitative coding process. We applied these methods to evaluate several ChatGPT models, including 3.5 Turbo, 4, 4o, o1-mini, and o1-preview. Our findings show that when used as problem solvers, LLMs generate correct final answers for 85.5% of the college algebra problems tested. When employed interactively as tutors, 90% of LLM dialogues show high-quality instructional support; however, many contain errors -- only 56.6% are entirely correct. We conclude that, despite their potential, LLMs are not yet suitable as intelligent tutors for math without human oversight or additional mechanisms to ensure correctness and quality.

Nicki Barari, Xin Lian, Christopher J. MacLellan

Deep neural networks have excelled in machine learning, particularly in vision tasks, however, they often suffer from catastrophic forgetting when learning new tasks sequentially. In this work, we introduce Cobweb4V, an alternative to traditional neural network approaches. Cobweb4V is a novel visual classification method that builds on Cobweb, a human like learning system that is inspired by the way humans incrementally learn new concepts over time. In this research, we conduct a comprehensive evaluation, showcasing Cobweb4Vs proficiency in learning visual concepts, requiring less data to achieve effective learning outcomes compared to traditional methods, maintaining stable performance over time, and achieving commendable asymptotic behavior, without catastrophic forgetting effects. These characteristics align with learning strategies in human cognition, positioning Cobweb4V as a promising alternative to neural network approaches.

Glen Smith, Adit Gupta, Christopher MacLellan

Intelligent tutoring systems (ITS) are effective for improving students' learning outcomes. However, their development is often complex, time-consuming, and requires specialized programming and tutor design knowledge, thus hindering their widespread application and personalization. We present the Apprentice Tutor Builder (ATB) , a platform that simplifies tutor creation and personalization. Instructors can utilize ATB's drag-and-drop tool to build tutor interfaces. Instructors can then interactively train the tutors' underlying AI agent to produce expert models that can solve problems. Training is achieved via using multiple interaction modalities including demonstrations, feedback, and user labels. We conducted a user study with 14 instructors to evaluate the effectiveness of ATB's design with end users. We found that users enjoyed the flexibility of the interface builder and ease and speed of agent teaching, but often desired additional time-saving features. With these insights, we identified a set of design recommendations for our platform and others that utilize interactive AI agents for tutor creation and customization.

Xin Lian, Sashank Varma, Christopher J. MacLellan

Cobweb, a human-like category learning system, differs from most cognitive science models in incrementally constructing hierarchically organized tree-like structures guided by the category utility measure. Prior studies have shown that Cobweb can capture psychological effects such as basic-level, typicality, and fan effects. However, a broader evaluation of Cobweb as a model of human categorization remains lacking. The current study addresses this gap. It establishes Cobweb's alignment with classical human category learning effects. It also explores Cobweb's flexibility to exhibit both exemplar- and prototype-like learning within a single framework. These findings set the stage for further research on Cobweb as a robust model of human category learning.

Daniel Weitekamp, Momin N. Siddiqui, Christopher J. MacLellan

Recent improvements in large language model (LLM) performance on academic benchmarks, such as MATH and GSM8K, have emboldened their use as standalone tutors and as simulations of human learning. However, these new applications require more than evaluations of final solution generation. We introduce TutorGym to evaluate these applications more directly. TutorGym is a standard interface for testing artificial intelligence (AI) agents within existing intelligent tutoring systems (ITS) that have been tested and refined in classroom studies, including Cognitive Tutors (CTAT), Apprentice Tutors, and OATutors. TutorGym is more than a simple problem-solution benchmark, it situates AI agents within the interactive interfaces of existing ITSs. At each step of problem-solving, AI agents are asked what they would do as a tutor or as a learner. As tutors, AI agents are prompted to provide tutoring support -- such as generating examples, hints, and step-level correctness feedback -- which can be evaluated directly against the adaptive step-by-step support provided by existing ITSs. As students, agents directly learn from ITS instruction, and their mistakes and learning trajectories can be compared to student data. TutorGym establishes a common framework for training and evaluating diverse AI agents, including LLMs, computational models of learning, and reinforcement learning agents, within a growing suite of learning environments. Currently, TutorGym includes 223 different tutor domains. In an initial evaluation, we find that current LLMs are poor at tutoring -- none did better than chance at labeling incorrect actions, and next-step actions were correct only ~52-70% of the time -- but they could produce remarkably human-like learning curves when trained as students with in-context learning.

Christopher J. MacLellan

Instructional designers face an overwhelming array of design choices, making it challenging to identify the most effective interventions. To address this issue, I propose the concept of a Model Human Learner, a unified computational model of learning that can aid designers in evaluating candidate interventions. This paper presents the first successful demonstration of this concept, showing that a computational model can accurately predict the outcomes of two human A/B experiments -- one testing a problem sequencing intervention and the other testing an item design intervention. It also demonstrates that such a model can generate learning curves without requiring human data and provide theoretical insights into why an instructional intervention is effective. These findings lay the groundwork for future Model Human Learners that integrate cognitive and learning theories to support instructional design across diverse tasks and interventions.

Darryl Hannan, Steven C. Nesbit, Ximing Wen et al.

Detecting elevated intracranial pressure (ICP) is crucial in diagnosing and managing various neurological conditions. These fluctuations in pressure are transmitted to the optic nerve sheath (ONS), resulting in changes to its diameter, which can then be detected using ultrasound imaging devices. However, interpreting sonographic images of the ONS can be challenging. In this work, we propose two systems that actively monitor the ONS diameter throughout an ultrasound video and make a final prediction as to whether ICP is elevated. To construct our systems, we leverage subject matter expert (SME) guidance, structuring our processing pipeline according to their collection procedure, while also prioritizing interpretability and computational efficiency. We conduct a number of experiments, demonstrating that our proposed systems are able to outperform various baselines. One of our SMEs then manually validates our top system's performance, lending further credibility to our approach while demonstrating its potential utility in a clinical setting.

Daniel Weitekamp, Christopher MacLellan, Erik Harpstead et al.

Human learning relies on specialization -- distinct cognitive mechanisms working together to enable rapid learning. In contrast, most modern neural networks rely on a single mechanism: gradient descent over an objective function. This raises the question: might human learners' relatively rapid learning from just tens of examples instead of tens of thousands in data-driven deep learning arise from our ability to use multiple specialized mechanisms of learning in combination? We investigate this question through an ablation analysis of inductive human learning simulations in online tutoring environments. Comparing reinforcement learning to a more data-efficient 3-mechanism symbolic rule induction approach, we find that decomposing learning into multiple distinct mechanisms significantly improves data efficiency, bringing it in line with human learning. Furthermore, we show that this decomposition has a greater impact on efficiency than the distinction between symbolic and subsymbolic learning alone. Efforts to align data-driven machine learning with human learning often overlook the stark difference in learning efficiency. Our findings suggest that integrating multiple specialized learning mechanisms may be key to bridging this gap.

Lane Lawley, Christopher J. MacLellan

We present a system for interpretable, symbolic, interactive task learning from dialog using a GPT model as a conversational front-end. The learned tasks are represented as hierarchical decompositions of predicate-argument structures with scoped variable arguments. By using a GPT model to convert interactive dialog into a semantic representation, and then recursively asking for definitions of unknown steps, we show that hierarchical task knowledge can be acquired and re-used in a natural and unrestrained conversational environment. We compare our system to a similar architecture using a more conventional parser and show that our system tolerates a much wider variety of linguistic variance.

Christopher J. MacLellan, Harshil Thakur

This paper presents a new concept formation approach that supports the ability to incrementally learn and predict labels for visual images. This work integrates the idea of convolutional image processing, from computer vision research, with a concept formation approach that is based on psychological studies of how humans incrementally form and use concepts. We experimentally evaluate this new approach by applying it to an incremental variation of the MNIST digit recognition task. We compare its performance to Cobweb, a concept formation approach that does not support convolutional processing, as well as two convolutional neural networks that vary in the complexity of their convolutional processing. This work represents a first step towards unifying modern computer vision ideas with classical concept formation research.

Daniel Weitekamp, Christopher MacLellan, Erik Harpstead et al.

Recent advances in machine learning have made it possible to train artificially intelligent agents that perform with super-human accuracy on a great diversity of complex tasks. However, the process of training these capabilities often necessitates millions of annotated examples -- far more than humans typically need in order to achieve a passing level of mastery on similar tasks. Thus, while contemporary methods in machine learning can produce agents that exhibit super-human performance, their rate of learning per opportunity in many domains is decidedly lower than human-learning. In this work we formalize a theory of Decomposed Inductive Procedure Learning (DIPL) that outlines how different forms of inductive symbolic learning can be used in combination to build agents that learn educationally relevant tasks such as mathematical, and scientific procedures, at a rate similar to human learners. We motivate the construction of this theory along Marr's concepts of the computational, algorithmic, and implementation levels of cognitive modeling, and outline at the computational-level six learning capacities that must be achieved to accurately model human learning. We demonstrate that agents built along the DIPL theory are amenable to satisfying these capacities, and demonstrate, both empirically and theoretically, that DIPL enables the creation of agents that exhibit human-like learning performance.

Devendra Singh Chaplot, Christopher MacLellan, Ruslan Salakhutdinov et al.

A cognitive model of human learning provides information about skills a learner must acquire to perform accurately in a task domain. Cognitive models of learning are not only of scientific interest, but are also valuable in adaptive online tutoring systems. A more accurate model yields more effective tutoring through better instructional decisions. Prior methods of automated cognitive model discovery have typically focused on well-structured domains, relied on student performance data or involved substantial human knowledge engineering. In this paper, we propose Cognitive Representation Learner (CogRL), a novel framework to learn accurate cognitive models in ill-structured domains with no data and little to no human knowledge engineering. Our contribution is two-fold: firstly, we show that representations learnt using CogRL can be used for accurate automatic cognitive model discovery without using any student performance data in several ill-structured domains: Rumble Blocks, Chinese Character, and Article Selection. This is especially effective and useful in domains where an accurate human-authored cognitive model is unavailable or authoring a cognitive model is difficult. Secondly, for domains where a cognitive model is available, we show that representations learned through CogRL can be used to get accurate estimates of skill difficulty and learning rate parameters without using any student performance data. These estimates are shown to highly correlate with estimates using student performance data on an Article Selection dataset.