Elena L. Glassman

Ian Arawjo, Chelse Swoopes, Priyan Vaithilingam et al. · harvard

Evaluating outputs of large language models (LLMs) is challenging, requiring making -- and making sense of -- many responses. Yet tools that go beyond basic prompting tend to require knowledge of programming APIs, focus on narrow domains, or are closed-source. We present ChainForge, an open-source visual toolkit for prompt engineering and on-demand hypothesis testing of text generation LLMs. ChainForge provides a graphical interface for comparison of responses across models and prompt variations. Our system was designed to support three tasks: model selection, prompt template design, and hypothesis testing (e.g., auditing). We released ChainForge early in its development and iterated on its design with academics and online users. Through in-lab and interview studies, we find that a range of people could use ChainForge to investigate hypotheses that matter to them, including in real-world settings. We identify three modes of prompt engineering and LLM hypothesis testing: opportunistic exploration, limited evaluation, and iterative refinement.

Joyce Zhou, Elena Glassman, Daniel S. Weld · harvard, uw

Scientists and science journalists, among others, often need to make sense of a large number of papers and how they compare with each other in scope, focus, findings, or any other important factors. However, with a large corpus of papers, it's cognitively demanding to pairwise compare and contrast them all with each other. Fully automating this review process would be infeasible, because it often requires domain-specific knowledge, as well as understanding what the context and motivations for the review are. While there are existing tools to help with the process of organizing and annotating papers for literature reviews, at the core they still rely on people to serially read through papers and manually make sense of relevant information. We present AVTALER, which combines peoples' unique skills, contextual awareness, and knowledge, together with the strength of automation. Given a set of comparable text excerpts from a paper corpus, it supports users in sensemaking and contrasting paper attributes by interactively aligning text excerpts in a table so that comparable details are presented in a shared column. AVTALER is based on a core alignment algorithm that makes use of modern NLP tools. Furthermore, AVTALER is a mixed-initiative system: users can interactively give the system constraints which are integrated into the alignment construction process.

Safinah Ali, Sohini Upadhyay, Gaurush Hiranandani et al. · harvard

Metric Elicitation (ME) is a framework for eliciting classification metrics that better align with implicit user preferences based on the task and context. The existing ME strategy so far is based on the assumption that users can most easily provide preference feedback over classifier statistics such as confusion matrices. This work examines ME, by providing a first ever implementation of the ME strategy. Specifically, we create a web-based ME interface and conduct a user study that elicits users' preferred metrics in a binary classification setting. We discuss the study findings and present guidelines for future research in this direction.

Simret Araya Gebreegziabher, Kuangshi Ai, Zheng Zhang et al.

Active Learning (AL) allows models to learn interactively from user feedback. This paper introduces a counterfactual data augmentation approach to AL, particularly addressing the selection of datapoints for user querying, a pivotal concern in enhancing data efficiency. Our approach is inspired by Variation Theory, a theory of human concept learning that emphasizes the essential features of a concept by focusing on what stays the same and what changes. Instead of just querying with existing datapoints, our approach synthesizes artificial datapoints that highlight potential key similarities and differences among labels using a neuro-symbolic pipeline combining large language models (LLMs) and rule-based models. Through an experiment in the example domain of text classification, we show that our approach achieves significantly higher performance when there are fewer annotated data. As the annotated training data gets larger the impact of the generated data starts to diminish showing its capability to address the cold start problem in AL. This research sheds light on integrating theories of human learning into the optimization of AL.

Alice Cai, Caine Ardayfio, AnhPhu Nguyen et al.

As technological advancements in extended reality (XR) amplify the demand for more XR content, traditional development processes face several challenges: 1) a steep learning curve for inexperienced developers, 2) a disconnect between 2D development environments and 3D user experiences inside headsets, and 3) slow iteration cycles due to context switching between development and testing environments. To address these challenges, we introduce AtomXR, a streamlined, immersive, no-code XR prototyping tool designed to empower both experienced and inexperienced developers in creating applications using natural language, eye-gaze, and touch interactions. AtomXR consists of: 1) AtomScript, a high-level human-interpretable scripting language for rapid prototyping, 2) a natural language interface that integrates LLMs and multimodal inputs for AtomScript generation, and 3) an immersive in-headset authoring environment. Empirical evaluation through two user studies offers insights into natural language-based and immersive prototyping, and shows AtomXR provides significant improvements in speed and user experience compared to traditional systems.

Celeste Barnaby, Koushik Sen, Tianyi Zhang et al.

Modern software engineering often involves using many existing APIs, both open source and, in industrial coding environments, proprietary. Programmers reference documentation and code search tools to remind themselves of proper common usage patterns of APIs. However, high-quality API usage examples are computationally expensive to curate and maintain, and API usage examples retrieved from company-wide code search can be tedious to review. We present a tool, EG, that mines codebases and shows the common, idiomatic usage examples for API methods. EG was integrated into Facebook's internal code search tool for the Hack language and evaluated on open-source GitHub projects written in Python. EG was also compared against code search results and hand-written examples from a popular programming website called ProgramCreek. Compared with these two baselines, examples generated by EG are more succinct and representative with less extraneous statements. In addition, a survey with Facebook developers shows that EG examples are preferred in 97 percent of cases.

Priyan Vaithilingam, Ian Arawjo, Elena L. Glassman

We ideate a future design workflow that involves AI technology. Drawing from activity and communication theory, we attempt to isolate the new value large AI models can provide design compared to past technologies. We arrive at three affordances -- dynamic grounding, constructive negotiation, and sustainable motivation -- that summarize latent qualities of natural language-enabled foundation models that, if explicitly designed for, can support the process of design. Through design fiction, we then imagine a future interface as a diegetic prototype, the story of Squirrel Game, that demonstrates each of our three affordances in a realistic usage scenario. Our design process, terminology, and diagrams aim to contribute to future discussions about the relative affordances of AI technology with regard to collaborating with human designers.

Chen Zhu-Tian, Zeyu Xiong, Xiaoshuo Yao et al.

Crafting effective prompts for code generation or editing with Large Language Models (LLMs) is not an easy task. Particularly, the absence of immediate, stable feedback during prompt crafting hinders effective interaction, as users are left to mentally imagine possible outcomes until the code is generated. In response, we introduce Language-Oriented Code Sketching, an interactive approach that provides instant, incremental feedback in the form of code sketches (i.e., incomplete code outlines) during prompt crafting. This approach converts a prompt into a code sketch by leveraging the inherent linguistic structures within the prompt and applying classic natural language processing techniques. The sketch then serves as an intermediate placeholder that not only previews the intended code structure but also guides the LLM towards the desired code, thereby enhancing human-LLM interaction. We conclude by discussing the approach's applicability and future plans.

Hai Dang, Chelse Swoopes, Daniel Buschek et al. · harvard

Many communities, including the scientific community, develop implicit writing norms. Understanding them is crucial for effective communication with that community. Writers gradually develop an implicit understanding of norms by reading papers and receiving feedback on their writing. However, it is difficult to both externalize this knowledge and apply it to one's own writing. We propose two new writing support concepts that reify document and sentence-level patterns in a given text corpus: (1) an ordered distribution over section titles and (2) given the user's draft and cursor location, many retrieved contextually relevant sentences. Recurring words in the latter are algorithmically highlighted to help users see any emergent norms. Study results (N=16) show that participants revised the structure and content using these concepts, gaining confidence in aligning with or breaking norms after reviewing many examples. These results demonstrate the value of reifying distributions over other authors' writing choices during the writing process.

Alice Cai, Ian Arawjo, Elena L. Glassman

The vast majority of discourse around AI development assumes that subservient, "moral" models aligned with "human values" are universally beneficial -- in short, that good AI is sycophantic AI. We explore the shadow of the sycophantic paradigm, a design space we term antagonistic AI: AI systems that are disagreeable, rude, interrupting, confrontational, challenging, etc. -- embedding opposite behaviors or values. Far from being "bad" or "immoral," we consider whether antagonistic AI systems may sometimes have benefits to users, such as forcing users to confront their assumptions, build resilience, or develop healthier relational boundaries. Drawing from formative explorations and a speculative design workshop where participants designed fictional AI technologies that employ antagonism, we lay out a design space for antagonistic AI, articulating potential benefits, design techniques, and methods of embedding antagonistic elements into user experience. Finally, we discuss the many ethical challenges of this space and identify three dimensions for the responsible design of antagonistic AI -- consent, context, and framing.

Priyan Vaithilingam, Munyeong Kim, Frida-Cecilia Acosta-Parenteau et al.

How do we update AI memory of user intent as intent changes? We consider how an AI interface may assist the integration of new information into a repository of natural language data. Inspired by software engineering concepts like impact analysis, we develop methods and a UI for managing semantic changes with non-local effects, which we call "semantic conflict resolution." The user commits new intent to a project -- makes a "semantic commit" -- and the AI helps the user detect and resolve semantic conflicts within a store of existing information representing their intent (an "intent specification"). We develop an interface, SemanticCommit, to better understand how users resolve conflicts when updating intent specifications such as Cursor Rules and game design documents. A knowledge graph-based RAG pipeline drives conflict detection, while LLMs assist in suggesting resolutions. We evaluate our technique on an initial benchmark. Then, we report a 12 user within-subjects study of SemanticCommit for two task domains -- game design documents, and AI agent memory in the style of ChatGPT memories -- where users integrated new information into an existing list. Half of our participants adopted a workflow of impact analysis, where they would first flag conflicts without AI revisions then resolve conflicts locally, despite having access to a global revision feature. We argue that AI agent interfaces, such as software IDEs like Cursor and Windsurf, should provide affordances for impact analysis and help users validate AI retrieval independently from generation. Our work speaks to how AI agent designers should think about updating memory as a process that involves human feedback and decision-making.

Lev Tankelevitch, Elena L. Glassman, Jessica He et al. · microsoft-research

Generative AI (GenAI) radically expands the scope and capability of automation for work, education, and everyday tasks, a transformation posing both risks and opportunities for human cognition. How will human cognition change, and what opportunities are there for GenAI to augment it? Which theories, metrics, and other tools are needed to address these questions? The CHI 2025 workshop on Tools for Thought aimed to bridge an emerging science of how the use of GenAI affects human thought, from metacognition to critical thinking, memory, and creativity, with an emerging design practice for building GenAI tools that both protect and augment human thought. Fifty-six researchers, designers, and thinkers from across disciplines as well as industry and academia, along with 34 papers and portfolios, seeded a day of discussion, ideation, and community-building. We synthesize this material here to begin mapping the space of research and design opportunities and to catalyze a multidisciplinary community around this pressing area of research.

Yewen Pu, Saujas Vaduguru, Priyan Vaithilingam et al.

The usage of Rational Speech Acts (RSA) framework has been successful in building \emph{pragmatic} program synthesizers that return programs which, in addition to being logically consistent with user-generated examples, account for the fact that a user chooses their examples informatively. We present a general method of amortizing the slow, exact RSA synthesizer. Our method first query the exact RSA synthesizer to compile a communication dataset. The dataset contains a number of example-dependent rankings of subsets of programs. It then distills a \textit{single} global ranking of all programs as an approximation to every ranking in the dataset. This global ranking is then used at inference time to rank multiple logically consistent candidate programs generated from a fast, non-pragmatic synthesizer. Experiments on two program synthesis domains using our ranking method resulted in orders of magnitudes of speed ups compared to the exact RSA synthesizer, while being more accurate than a non-pragmatic synthesizer when communicating with humans. Finally, we prove that in the special case of synthesis from a single example, this approximation is exact.

Katy Ilonka Gero, Chelse Swoopes, Ziwei Gu et al.

Large language models (LLMs) are capable of generating multiple responses to a single prompt, yet little effort has been expended to help end-users or system designers make use of this capability. In this paper, we explore how to present many LLM responses at once. We design five features, which include both pre-existing and novel methods for computing similarities and differences across textual documents, as well as how to render their outputs. We report on a controlled user study (n=24) and eight case studies evaluating these features and how they support users in different tasks. We find that the features support a wide variety of sensemaking tasks and even make tasks previously considered to be too difficult by our participants now tractable. Finally, we present design guidelines to inform future explorations of new LLM interfaces.

Yewen Pu, Saujas Vaduguru, Priyan Vaithilingam et al.

In program synthesis, an intelligent system takes in a set of user-generated examples and returns a program that is logically consistent with these examples. The usage of Rational Speech Acts (RSA) framework has been successful in building \emph{pragmatic} program synthesizers that return programs which -- in addition to being logically consistent -- account for the fact that a user chooses their examples informatively. However, the computational burden of running the RSA algorithm has restricted the application of pragmatic program synthesis to domains with a small number of possible programs. This work presents a novel method of amortizing the RSA algorithm by leveraging a \emph{global pragmatic ranking} -- a single, total ordering of all the hypotheses. We prove that for a pragmatic synthesizer that uses a single demonstration, our global ranking method exactly replicates RSA's ranked responses. We further empirically show that global rankings effectively approximate the full pragmatic synthesizer in an online, multi-demonstration setting. Experiments on two program synthesis domains using our pragmatic ranking method resulted in orders of magnitudes of speed ups compared to the RSA synthesizer, while outperforming the standard, non-pragmatic synthesizer.

Mina Khan, Zeel Patel, Kathryn Wantlin et al.

Technology based screentime, the time an individual spends engaging with their computer or cell phone, has increased exponentially over the past decade, but perhaps most alarmingly amidst the COVID-19 pandemic. Although many software based interventions exist to reduce screentime, users report a variety of issues relating to the timing of the intervention, the strictness of the tool, and its ability to encourage organic, long-term habit formation. We develop guidelines for the design of behaviour intervention software by conducting a survey to investigate three research questions and further inform the mechanisms of computer-related behaviour change applications. RQ1: What do people want to change and why/how? RQ2: What applications do people use or have used, why do they work or not, and what additional support is desired? RQ3: What are helpful/unhelpful computer breaks and why? Our survey had 68 participants and three key findings. First, time management is a primary concern, but emotional and physical side-effects are equally important. Second, site blockers, self-trackers, and timers are commonly used, but they are ineffective as they are easy-to-ignore and not personalized. Third, away-from-computer breaks, especially involving physical activity, are helpful, whereas on-screen breaks are unhelpful, especially when they are long, because they are not refreshing. We recommend personalized and closed-loop computer-usage behaviour change support and especially encouraging off-the-computer screentime breaks.

Ziv Epstein, Nicolò Foppiani, Sophie Hilgard et al.

Social media platforms are increasingly deploying complex interventions to help users detect false news. Labeling false news using techniques that combine crowd-sourcing with artificial intelligence (AI) offers a promising way to inform users about potentially low-quality information without censoring content, but also can be hard for users to understand. In this study, we examine how users respond in their sharing intentions to information they are provided about a hypothetical human-AI hybrid system. We ask i) if these warnings increase discernment in social media sharing intentions and ii) if explaining how the labeling system works can boost the effectiveness of the warnings. To do so, we conduct a study ($N=1473$ Americans) in which participants indicated their likelihood of sharing content. Participants were randomly assigned to a control, a treatment where false content was labeled, or a treatment where the warning labels came with an explanation of how they were generated. We find clear evidence that both treatments increase sharing discernment, and directional evidence that explanations increase the warnings' effectiveness. Interestingly, we do not find that the explanations increase self-reported trust in the warning labels, although we do find some evidence that participants found the warnings with the explanations to be more informative. Together, these results have important implications for designing and deploying transparent misinformation warning labels, and AI-mediated systems more broadly.

Andrew Slavin Ross, Nina Chen, Elisa Zhao Hang et al.

For machine learning models to be most useful in numerous sociotechnical systems, many have argued that they must be human-interpretable. However, despite increasing interest in interpretability, there remains no firm consensus on how to measure it. This is especially true in representation learning, where interpretability research has focused on "disentanglement" measures only applicable to synthetic datasets and not grounded in human factors. We introduce a task to quantify the human-interpretability of generative model representations, where users interactively modify representations to reconstruct target instances. On synthetic datasets, we find performance on this task much more reliably differentiates entangled and disentangled models than baseline approaches. On a real dataset, we find it differentiates between representation learning methods widely believed but never shown to produce more or less interpretable models. In both cases, we ran small-scale think-aloud studies and large-scale experiments on Amazon Mechanical Turk to confirm that our qualitative and quantitative results agreed.

José Pablo Cambronero, Jiasi Shen, Jürgen Cito et al.

We present a study that characterizes the way developers use automatically generated patches when fixing software defects. Our study tasked two groups of developers with repairing defects in C programs. Both groups were provided with the defective line of code. One was also provided with five automatically generated and validated patches, all of which modified the defective line of code, and one of which was correct. Contrary to our initial expectations, the group with access to the generated patches did not produce more correct patches and did not produce patches in less time. We characterize the main behaviors observed in experimental subjects: a focus on understanding the defect and the relationship of the patches to the original source code. Based on this characterization, we highlight various potentially productive directions for future developer-centric automatic patch generation systems.

Ryo Suzuki, Gustavo Soares, Andrew Head et al.

Recent advances in program synthesis offer means to automatically debug student submissions and generate personalized feedback in massive programming classrooms. When automatically generating feedback for programming assignments, a key challenge is designing pedagogically useful hints that are as effective as the manual feedback given by teachers. Through an analysis of teachers' hint-giving practices in 132 online Q&A posts, we establish three design guidelines that an effective feedback design should follow. Based on these guidelines, we develop a feedback system that leverages both program synthesis and visualization techniques. Our system compares the dynamic code execution of both incorrect and fixed code and highlights how the error leads to a difference in behavior and where the incorrect code trace diverges from the expected solution. Results from our study suggest that our system enables students to detect and fix bugs that are not caught by students using another existing visual debugging tool.

Joseph Jay Williams, Korinn Ostrow, Xiaolu Xiong et al.

In contrast to typical laboratory experiments, the everyday use of online educational resources by large populations and the prevalence of software infrastructure for A/B testing leads us to consider how platforms can embed in vivo experiments that do not merely support research, but ensure practical improvements to their educational components. Examples are presented of randomized experimental comparisons conducted by subsets of the authors in three widely used online educational platforms Khan Academy, edX, and ASSISTments. We suggest design principles for platform technology to support randomized experiments that lead to practical improvements enabling Iterative Improvement and Collaborative Work and explain the benefit of their implementation by WPI co-authors in the ASSISTments platform.