Pith. sign in

REVIEW 4 major objections 6 minor 115 references

GLOSS: Group of LLMs for Open-Ended Sensemaking of Passive Sensing Data for Health and Wellbeing

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that GLOSS, a group of collaborating LLM agents that generate and execute code over raw sensor databases, can answer open-ended health and wellbeing questions from passive sensing data far more accurately than standard…

desk verdict Interesting system, but the headline accuracy numbers rest on a self-scored evaluation that needs independent verification before they can be trusted. read the letter →

arxiv 2507.05461 v2 pith:UPVMFPM4 submitted 2025-07-07 cs.HC

Akshat Choube , Ha Le , Jiachen Li , Kaixin Ji , Vedant Das Swain , Varun Mishra This is my paper
classification cs.HC
keywords passivesensingsensemakinglargelanguagemodelsretrieval-augmentedgenerationmobilehealthpersonalinformaticscodemultimodaldatatriangulation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

GLOSS is a proposed sensemaking system that lets people ask natural-language questions about raw smartphone and wearable sensor data and get answers that require combining multiple data streams, such as "did this user exercise outdoors?" or "why might this user have reported stress on Monday evening?" The paper argues that this task needs more than retrieval: answers often require computation, comparison, and triangulation across location, activity, app usage, heart rate, and other streams. To do that, GLOSS organizes several LLM agents into an information-seeking loop and a sensemaking loop, where one agent writes and executes code against raw databases and another converts results into natural language, with a running "understanding" refined until the query is answered. The paper's load-bearing evaluation result is that GLOSS answers objective and mixed queries with 87.93% accuracy versus 29.31% for a standard RAG baseline, and with significantly higher consistency. If true, this would make deep sensor-data interpretation accessible to clinicians, behavioral scientists, and self-trackers without programming expertise.

What carries the argument

The load-bearing machinery is the two-loop agent architecture with a code-generation core rather than a single retrieval circuit. In the information-seeking loop, an agent that generates executable Python or bash code runs inside a sandbox, calls predefined helper functions that query raw per-stream databases (location, activity, app usage, steps, Wi-Fi, calls, battery, heart rate, and a stress model), and then triangulates across streams on demand. Because code, not the language model, does the arithmetic and joins, the system can answer computation-heavy queries accurately. In the sensemaking loop, a local agent converts raw outputs into natural language, a global agent maintains a cumulative "understanding" and notes what additional data is needed, and a next-step agent checks whether the understanding answers the query, with a five-iteration cutoff to avoid infinite loops. The final answer is formatted by a presentation agent that can tailor output to human-readable or machine-readable needs.

What would settle it

Have two independent researchers, blind to GLOSS's and RAG's outputs, write ground-truth code for the same 60 queries or a fresh sample and compare; if the independently computed accuracy for GLOSS does not remain far above RAG's, the claimed 87.93% versus 29.31% margin is not reproducible.

Watch

Extended reading notes

Core claim

The central claim is that open-ended sensemaking over passive sensing data can be automated by a group of cooperating LLM agents that fetch raw data, generate and execute code to process it, and iteratively build an understanding, rather than by retrieving pre-chunked data and generating text. GLOSS is the instantiation: an Action Plan agent decides whether and how the query can be answered, an Information Seeking agent formulates data requests, a Database/Model Manager selects retrieval helper functions, a Code Generation agent writes and runs Python or bash code that can triangulate multiple sensor streams, a Local Sensemaking agent renders results in natural language, a Global Sensemaking agent updates the accumulated understanding, a Next Step agent decides when to stop, and a Presentation agent formats the answer. On a crowd-sourced set of 244 objective, subjective, and mixed queries (with two participants per original query), GLOSS reports 87.93% accuracy on a random 60-query sample and 66.19% consistency across 210 queries, versus RAG's 29.31% accuracy and 52.85% consistency, with the accuracy and consistency differences reported as statistically significant. The authors attribute the gap to GLOSS performing calculations and triangulation in code rather than relying on the LLM's arithmetic, and they note that RAG frequently produced plausible but made-up numbers.

Load-bearing premise

The evaluation's ground truth for the 60 sampled objective and mixed queries was written by the first author alone as hand-written Python code, so the central accuracy comparison rests on that code being correct and unbiased.

Editorial extensions

If this is right

  • For systems that answer health queries from raw sensor streams, code-generating agents over raw data should be preferred over text-retrieval RAG whenever queries involve arithmetic, thresholds, or joins across streams.
  • Non-programming stakeholders, including clinicians, behavioral scientists, and self-trackers, could run open-ended analyses that previously required a data scientist or a custom dashboard.
  • The intermediate action plans, memory, and understanding shown to users give GLOSS a transparency advantage over end-to-end LLM answers, supporting trust in research and clinical settings.
  • The same core system can be embedded in automated pipelines: generating context-aware EMA questions, adding interpretability to black-box anomaly detectors, and producing personalized reflection narratives.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because the 60-query accuracy sample was ground-truthed by a single annotator, an independent replication with pre-registered ground-truth code is the natural next step before treating 87.93% as a stable number.
  • The 66.19% consistency suggests a concrete improvement the paper mentions but does not test: inserting a clarification step for ambiguous queries could resolve cases where GLOSS answers correctly via different valid paths.
  • The RAG baseline may be conservative, since it retrieves text chunks of raw data; a stronger RAG variant with schema-aware retrieval or precomputed aggregates might shrink the gap, and that comparison is not yet in the literature.
  • If the result generalizes, the same agent-group pattern could be applied to other high-dimensional personal data, including images and audio, as the paper itself anticipates.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper introduces GLOSS, a multi-agent LLM system for open-ended sensemaking of raw passive sensing data. GLOSS combines action-plan generation, an information-seeking/sensemaking loop, code generation over raw sensor databases, and a presentation agent. The authors evaluate GLOSS against a RAG baseline on 244 crowd-sourced queries, reporting that GLOSS achieves 87.93% accuracy and 66.19% consistency on objective/mixed queries versus RAG's 29.31% accuracy and 52.85% consistency, and they also report subjective ratings on relevance, interpretation, domain knowledge, logic, and clarity. Four use cases are presented: supporting non-CS researchers, generating reflection narratives, AI-triggered EMA prompting, and enhancing interpretability of black-box models.

Significance. If the central comparison is sound, GLOSS is a meaningful step toward making passive sensing data accessible to non-programmers, and the system design is well grounded in sensemaking theory. The paper ships public code and queries, which is a strength for reproducibility. However, the accuracy evaluation rests on ground truth written by the first author after seeing model outputs, and the RAG baseline is underspecified; these limitations directly affect the headline claim. The use cases are illustrative rather than fully evaluated, as the authors acknowledge, so the contribution should be judged primarily on the core system and its comparison.

major comments (4)
  1. [§4.2, Table 5] The accuracy ground truth is not independent: the first author wrote Python code 'based on the logic model followed' by GLOSS after the model runs. This can ratify plausible-but-wrong interpretations of ambiguous queries, such as defining 'mobility' as step count rather than GPS displacement. With only 60 queries and a reported gap of 58.6 percentage points, even a modest systematic scoring bias could materially change the conclusion. Please provide the scoring code, have at least one independent annotator re-derive ground truth for all 60 queries or a pre-registered subset, report inter-annotator agreement, and state how ambiguous queries were adjudicated.
  2. [§5.1, §3.2] Consistency is measured as 'identical across all three runs' using GPT-4o with temperature=1 and top_p=1. It is not clear whether 'identical' means verbatim string equality, normalized equality, or semantic equivalence. Under stochastic sampling, verbatim matching likely understates consistency for both systems. Please specify the matching criterion, report the distribution of consistency failures (e.g., changed wording versus changed numerical answer), and consider reporting semantic consistency with multiple samples.
  3. [§4.2] The RAG baseline is underspecified. The text states that raw sensor data were converted to natural language and stored in a Chroma database with LangChain's RAG framework, but it does not report chunk sizes, retrieval top-k, the prompt template, or whether retrieval was per sensor stream or over a combined corpus. Because the central claim is that GLOSS outperforms 'commonly used RAG,' the baseline needs enough detail to be reproduced and to be recognizable as a reasonable instantiation of RAG for this task.
  4. [§5.2] The subjective evaluation does not report the number of responses per condition, inter-coder reliability, or the full contingency tables behind the chi-square tests. The description that 'both response 1 and response 2 included a random number of GLOSS and RAG-generated responses' is ambiguous about whether each annotator saw exactly one response per query from each system. Please report Cohen's kappa, sample sizes, and collapsed category definitions; with 34 subjective queries, chi-square tests with df=1 may be fragile if categories were collapsed.
minor comments (6)
  1. [Title page / author block] The author block contains a typo: 'le.ha1@northeastern.edu,' has a stray comma, and Kaixin Ji's email is listed as 'li.jiachen4@northeastern.edu,' which appears to be a copy of Jiachen Li's email rather than a distinct address.
  2. [§1] The word 'anixety' should be 'anxiety' in the sentence 'passive mobile sensing has demonstrated significant potential in monitoring and assessing various health and well-being outcomes, including depression [27,99,104], stress [25,71,73,74,90,99], and anixety [96].'
  3. [§4.2] The phrase 'process & analyze the date to calculate accuracy' should read 'the data' rather than 'the date.'
  4. [§5.2.4] The heading 'Domain Knowlegde' is misspelled; it should be 'Domain Knowledge.'
  5. [Table 2] The legend for Table 2 is incomplete: the sentence fragment 'where indicates the presence of a particular capability in the given approach' is cut off and does not explain the symbol used in the 'Multi data streams' column.
  6. [Figure 4] Figure 4 shows subjective ratings without error bars or per-condition sample sizes; adding these would clarify the comparison and support the reported chi-square tests.

Circularity Check

1 steps flagged · score 6.0 of 10

The objective-query accuracy metric is self-referential: Section 4.2 defines accuracy as matching 'the underlying logic it followed' and builds the ground-truth code from that same logic, so the headline 87.93% vs. 29.31% gap is not an independent measurement.

  1. self definitional [Section 4.2, Evaluation Metrics]
    "Accuracy is measured by evaluating whether the model retrieved the correct answer based on the underlying logic it followed. ... Thus, to evaluate the accuracy of models, we needed to manually write Python code and analyze the data based on the logic model followed."

    The correctness criterion is defined by the very logic the system chose to follow. The first author writes the scoring code 'based on the logic model followed,' so any answer that faithfully executes the system's own interpretation (e.g., mobility as step count versus GPS displacement) is marked correct by construction rather than checked against an independent ground truth. The paper itself acknowledges that objective queries can have 'multiple valid logical paths' and that GLOSS 'followed different logical paths in different runs.' Consequently, the reported 87.93% accuracy measures internal agreement between the system's output and a re-implementation of that same logic, not agreement with an external answer.

full rationale

Most of this systems paper is not circular: there is no fitted equation, no derivation chain, and no imported uniqueness theorem. The four use cases are demonstrations, not predictions. The self-citations ([19], [61], [74]) are used for background, related work, and a stress ML model that is a data stream, not a load-bearing justification of the headline claim. The consistency metric is independently defined as identical responses across three runs, and the subjective evaluation uses blinded annotators with predefined criteria. However, the objective-query accuracy evaluation, which carries the central quantitative claim, is self-referential: Section 4.2 defines accuracy as whether the model retrieved the correct answer 'based on the underlying logic it followed' and then has the first author write the ground-truth code 'based on the logic model followed.' This means the scoring code is derived from the system's own chosen interpretation, so the accuracy number is partly guaranteed by construction rather than by independent verification. Because the headline comparison with RAG depends on this accuracy metric, the paper exhibits partial circularity in its central evaluation, though the subjective evaluation and consistency results retain independent content.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No numeric parameters were fitted to data in this paper. The only adjustable settings are LLM inference hyperparameters (temperature=1, top_p=1) and the five-iteration cutoff, which are implementation choices rather than fitted values. GLOSS introduces software agents as system components, not new physical or theoretical entities, so no invented entities are recorded.

assumptions (4)
  • domain assumption Converting code outputs to natural language preserves enough information for the global sensemaking agent to reason correctly.
    Section 3.1.2 describes the Local Sensemaking Agent converting Python data structures into natural language before global sensemaking. The correctness of downstream reasoning depends on this conversion not losing critical detail.
  • ad hoc to paper The five-iteration loop cutoff is sufficient for the sensemaking process to reach accurate answers.
    Section 3.2 states 'we have set up a hard cutoff of five iterations, after which GLOSS presents an answer to the user derived from the latest understanding.' No analysis shows convergence by five iterations.
  • domain assumption The RAG baseline is a representative implementation of prior retrieval-based passive-sensing systems.
    Section 4.2 says the baseline uses LangChain's RAG with GPT-4o over natural-language-transformed sensor data. The comparison's fairness depends on this implementation matching how prior work uses RAG.
  • domain assumption Manually coded ground truth by the first author is correct for the 60 sampled objective queries.
    Section 4.2 states 'the first author wrote code to process & analyze the data to calculate accuracy for these 60 queries.' No inter-rater reliability or independent verification is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GLOSS: Group of LLMs for Open-Ended Sensemaking of Passive Sensing Data for Health and Wellbeing." pith.science (2026). https://pith.science/paper/UPVMFPM4

@misc{pith2026250705461,
  author       = {Pith},
  title        = {Pith review of: GLOSS: Group of LLMs for Open-Ended Sensemaking of Passive Sensing Data for Health and Wellbeing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UPVMFPM4}},
  note         = {Machine review of arXiv:2507.05461}
}
read the original abstract

The ubiquitous presence of smartphones and wearables has enabled researchers to build prediction and detection models for various health and behavior outcomes using passive sensing data from these devices. Achieving a high-level, holistic understanding of an individual's behavior and context, however, remains a significant challenge. Due to the nature of passive sensing data, sensemaking -- the process of interpreting and extracting insights -- requires both domain knowledge and technical expertise, creating barriers for different stakeholders. Existing systems designed to support sensemaking are either not open-ended or cannot perform complex data triangulation. In this paper, we present a novel sensemaking system, Group of LLMs for Open-ended Sensemaking (GLOSS), capable of open-ended sensemaking and performing complex multimodal triangulation to derive insights. We demonstrate that GLOSS significantly outperforms the commonly used Retrieval-Augmented Generation (RAG) technique, achieving 87.93% accuracy and 66.19% consistency, compared to RAG's 29.31% accuracy and 52.85% consistency. Furthermore, we showcase the promise of GLOSS through four use cases inspired by prior and ongoing work in the UbiComp and HCI communities. Finally, we discuss the potential of GLOSS, its broader implications, and the limitations of our work.

Figures

Figures reproduced from arXiv: 2507.05461 by the authors.

Figure 1
Figure 1. GLOSS system design where different LLM agents interact according to prior sensemaking theories [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. (a) Action plan when query requests data not supported by GLOSS (b) Understanding when GLOSS has technical [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Example descriptions of databases (left) and helper functions (right) provided to agents for action plan and code [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Comparison of GLOSS and RAG across mul￾tiple dimensions. We found statistical evidence be￾tween the subjective measurements of GLOSS and RAG in terms of logic and interpretation (highlighted with *** on top of the bars) The trends in subjective evaluations were more su…
Figure 5
Figure 5. Figure 5: GLOSS UI design that shows relevant components of sensemaking (such as Action Plan, Information Requests, etc.). [PITH_FULL_IMAGE:figures/full_fig_p019_5.png]
Figure 6
Figure 6. Figure 6: GLOSS generates different narratives to the same query when instructed to focus on quantitative and qualitative [PITH_FULL_IMAGE:figures/full_fig_p021_6.png]
Figure 7
Figure 7. Figure 7: GLOSS can model the context using passive sensing data and then use the context to generate relevant questions to [PITH_FULL_IMAGE:figures/full_fig_p023_7.png]
Figure 8
Figure 8. Figure 8: Black box models to be augmented with GLOSS generated insights to enhance interpretability of results supporting [PITH_FULL_IMAGE:figures/full_fig_p024_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

115 extracted references · 67 canonical work pages

  1. [1]

    Sensing technologies for monitoring serious mental illnesses

    Saeed Abdullah and Tanzeem Choudhury. Sensing technologies for monitoring serious mental illnesses. IEEE MultiMedia, 25(1):61–75,

  2. [2]

    Adler, Yuewen Yang, Thalia Viranda, Xuhai Xu, David C

    Daniel A. Adler, Yuewen Yang, Thalia Viranda, Xuhai Xu, David C. Mohr, Anna R. Van Meter, Julia C. Tartaglia, Nicholas C. Jacobson, Fei Wang, Deborah Estrin, and Tanzeem Choudhury. Beyond detection: Towards actionable sensing research in clinical mental healthcare. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. , 8(4), November 2024. Number of page...

    2024

  3. [3]

    Mindful-RAG: A Study of Points of Failure in Retrieval Augmented Generation, October 2024

    Garima Agrawal, Tharindu Kumarage, Zeyad Alghamdi, and Huan Liu. Mindful-RAG: A Study of Points of Failure in Retrieval Augmented Generation, October 2024. arXiv:2407.12216 [cs]

    arXiv 2024

  4. [4]

    Mobile and wearable sensors for data-driven health monitoring system: State-of-the-art and future prospect

    Chioma Virginia Anikwe, Henry Friday Nweke, Anayo Chukwu Ikegwu, Chukwunonso Adolphus Egwuonwu, Fergus Uchenna Onu, Uzoma Rita Alo, and Ying Wah Teh. Mobile and wearable sensors for data-driven health monitoring system: State-of-the-art and future prospect. Expert Systems with Applications, 202:117362, 2022. Publisher: Elsevier

    2022

  5. [5]

    Frustrated and confused: the American public rates its cancer-related information-seeking experiences

    Neeraj K Arora, Bradford W Hesse, Barbara K Rimer, Kasisomayajula Viswanath, Marla L Clayman, and Robert T Croyle. Frustrated and confused: the American public rates its cancer-related information-seeking experiences. Journal of general internal medicine , 23:223–228, 2008. Publisher: Springer

    2008

  6. [6]

    Abandonment of personal quantification: A review and empirical study investigating reasons for wearable activity tracking attrition

    Christiane Attig and Thomas Franke. Abandonment of personal quantification: A review and empirical study investigating reasons for wearable activity tracking attrition. Computers in Human Behavior , 102:223–237, 2020. Publisher: Elsevier

    2020

  7. [7]

    Ling Bao and Stephen S. Intille. Activity recognition from user-annotated acceleration data. In Takeo Kanade, Josef Kittler, Jon M. Kleinberg, Friedemann Mattern, John C. Mitchell, Oscar Nierstrasz, C. Pandu Rangan, Bernhard Steffen, Demetri Terzopoulos, Dough , Vol. 1, No. 1, Article . Publication date: September 2025. GLOSS: Group of LLMs for Open-Ended...

    2025

  8. [8]

    The interpretation and exploitation of information in criminal investigations

    Emma Caroline Barrett. The interpretation and exploitation of information in criminal investigations . phd, University of Birmingham, 2009

    2009

Show all 115 references
  1. [9]

    Reviewing reflection: on the use of reflection in interactive system design

    Eric PS Baumer, Vera Khovanskaya, Mark Matthews, Lindsay Reynolds, Victoria Schwanda Sosik, and Geri Gay. Reviewing reflection: on the use of reflection in interactive system design. In Proceedings of the 2014 conference on Designing interactive systems , pages 93–102, 2014

    2014

  2. [10]

    Revisiting reflection in hci: Four design resources for technologies that support reflection

    Marit Bentvelzen, Paweł W Woźniak, Pia SF Herbes, Evropi Stefanidi, and Jasmin Niess. Revisiting reflection in hci: Four design resources for technologies that support reflection. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies , 6(1):1–27, ...

    2022

  3. [11]

    Detecting stress in college freshman from wearable sleep data

    Laura Bloomfield, Mikaela Fudoligb, Peter Doddsb, and Llorinb Jordan. Detecting stress in college freshman from wearable sleep data

  4. [12]

    Go and grow: Mapping personal data to a living plant

    Fadi Botros, Charles Perin, Bon Adriel Aseniero, and Sheelagh Carpendale. Go and grow: Mapping personal data to a living plant. In Proceedings of the international working conference on advanced visual interfaces , pages 112–119, 2016

    2016

  5. [13]

    Patient engagement in a multimodal digital phenotyping study of opioid use disorder

    Cynthia I Campbell, Ching-Hua Chen, Sara R Adams, Asma Asyyed, Ninad R Athale, Monique B Does, Saeed Hassanpour, Emily Hichborn, Melanie Jackson-Morris, Nicholas C Jacobson, and others. Patient engagement in a multimodal digital phenotyping study of opioid use disorder. Journa...

    2023

  6. [14]

    Challenges and recommendations for wearable devices in digital health: Data quality, interoperability, health equity, fairness

    Stefano Canali, Viola Schiaffonati, and Andrea Aliverti. Challenges and recommendations for wearable devices in digital health: Data quality, interoperability, health equity, fairness. PLOS Digital Health, 1(10):e0000104, 2022. Publisher: Public Library of Science San Francisc...

    2022

  7. [15]

    Methods in predictive techniques for mental health status on social media: a critical review

    Stevie Chancellor and Munmun De Choudhury. Methods in predictive techniques for mental health status on social media: a critical review. NPJ digital medicine, 3(1):43, 2020. Publisher: Nature Publishing Group UK London

    2020

  8. [16]

    Sensor2Text: Enabling natural language interactions for daily activity tracking using wearable sensors

    Wenqiang Chen, Jiaxuan Cheng, Leyao Wang, Wei Zhao, and Wojciech Matusik. Sensor2Text: Enabling natural language interactions for daily activity tracking using wearable sensors. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies , 8(4):1–26, 20...

    2024

  9. [17]

    Understanding client support strategies to improve clinical outcomes in an online mental health intervention

    Prerna Chikersal, Danielle Belgrave, Gavin Doherty, Angel Enrique, Jorge E Palacios, Derek Richards, and Anja Thieme. Understanding client support strategies to improve clinical outcomes in an online mental health intervention. In Proceedings of the 2020 CHI conference on huma...

    2020

  10. [18]

    Understanding quantified-selfers’ practices in collecting and exploring personal data

    Eun Kyoung Choe, Nicole B Lee, Bongshin Lee, Wanda Pratt, and Julie A Kientz. Understanding quantified-selfers’ practices in collecting and exploring personal data. In Proceedings of the SIGCHI conference on human factors in computing systems , pages 1143–1152, 2014

    2014

  11. [19]

    Imputation Matters: A Deeper Look into an Overlooked Step in Longitudinal Health and Behavior Sensing Research, December 2024

    Akshat Choube, Rahul Majethia, Sohini Bhattacharya, Vedant Das Swain, Jiachen Li, and Varun Mishra. Imputation Matters: A Deeper Look into an Overlooked Step in Longitudinal Health and Behavior Sensing Research, December 2024. arXiv:2412.06018 [stat]

    2024 arXiv

  12. [20]

    SeSaMe: a framework to simulate self-reported ground truth for mental health sensing studies

    Akshat Choube, Vedant Das Swain, and Varun Mishra. SeSaMe: a framework to simulate self-reported ground truth for mental health sensing studies. arXiv preprint arXiv:2403.17219, 2024

    2024 arXiv

  13. [21]

    Health information needs, sources, and barriers of primary care patients to achieve patient-centered care: A literature review

    Martina A Clarke, Joi L Moore, Linsey M Steege, Richelle J Koopman, Jeffery L Belden, Shannon M Canfield, Susan E Meadows, Susan G Elliott, and Min Soon Kim. Health information needs, sources, and barriers of primary care patients to achieve patient-centered care: A literature...

    2016

  14. [22]

    Grounded theory research: Procedures, canons, and evaluative criteria

    Juliet M Corbin and Anselm Strauss. Grounded theory research: Procedures, canons, and evaluative criteria. Qualitative sociology, 13(1):3–21, 1990. Publisher: Springer

    1990

  15. [23]

    Furlotte, Zhun Yang, Chace Lee, Erik Schenck, Yojan Patel, Jian Cui, Logan Douglas Schneider, Robby Bryant, Ryan G

    Justin Cosentino, Anastasiya Belyaeva, Xin Liu, Nicholas A. Furlotte, Zhun Yang, Chace Lee, Erik Schenck, Yojan Patel, Jian Cui, Logan Douglas Schneider, Robby Bryant, Ryan G. Gomes, Allen Jiang, Roy Lee, Yun Liu, Javier Perez, Jameson K. Rogers, Cathy Speed, Shyam Tailor, Meg...

    2024 arXiv

  16. [24]

    Data sensemaking in self-tracking: Towards a new generation of self-tracking tools

    Aykut Coşkun and Armağan Karahanoğlu. Data sensemaking in self-tracking: Towards a new generation of self-tracking tools. International Journal of Human–Computer Interaction , 39(12):2339–2360, 2023. Publisher: Taylor & Francis

    2023

  17. [25]

    Semantic gap in predicting mental wellbeing through passive sensing

    Vedant Das Swain, Victor Chen, Shrija Mishra, Stephen M Mattingly, Gregory D Abowd, and Munmun De Choudhury. Semantic gap in predicting mental wellbeing through passive sensing. In Proceedings of the 2022 CHI conference on human factors in computing systems , 2022

    2022

  18. [26]

    Benchmarks for automated commonsense reasoning: A survey

    Ernest Davis. Benchmarks for automated commonsense reasoning: A survey. ACM Computing Surveys, 56(4):1–41, 2023. Publisher: ACM New York, NY, USA

    2023

  19. [27]

    Predicting depression via social media

    Munmun De Choudhury, Michael Gamon, Scott Counts, and Eric Horvitz. Predicting depression via social media. Proceedings of the International AAAI Conference on Web and Social Media , 7(1):128–137, August 2021

    2021

  20. [28]

    From the mind’s eye of the “user”: The sense-making qualitative-quantitative methodology

    Brenda Dervin. From the mind’s eye of the “user”: The sense-making qualitative-quantitative methodology. Qualitative Research in Information Management/Libraries Unlimited, 1992. , Vol. 1, No. 1, Article . Publication date: September 2025. 28 • Akshat Choube, Ha Le, Jiachen Li...

    1992

  21. [29]

    An overview of sense-making research: Concepts, methods, and results to date

    Brenda Dervin and others. An overview of sense-making research: Concepts, methods, and results to date. 1983. Publisher: Author

    1983

  22. [30]

    The role of feedback in the process of health behavior change

    Carlo C DiClemente, Angela S Marinilli, Manu Singh, and Lori E Bellino. The role of feedback in the process of health behavior change. American journal of health behavior , 25(3):217–227, 2001. Publisher: PNG Publications and Scientific Research Limited

    2001

  23. [31]

    Deep neural networks in psychiatry

    Daniel Durstewitz, Georgia Koppe, and Andreas Meyer-Lindenberg. Deep neural networks in psychiatry. Molecular psychiatry, 24(11):1583–1598, 2019. Publisher: Nature Publishing Group UK London

    2019

  24. [32]

    Imputation of missing longitudinal data: a comparison of methods

    J Engels. Imputation of missing longitudinal data: a comparison of methods. Journal of Clinical Epidemiology , 56(10):968–976, October 2003

    2003

  25. [33]

    Epstein, Clara Caldeira, Mayara Costa Figueiredo, Xi Lu, Lucas M

    Daniel A. Epstein, Clara Caldeira, Mayara Costa Figueiredo, Xi Lu, Lucas M. Silva, Lucretia Williams, Jong Ho Lee, Qingyang Li, Simran Ahuja, Qiuer Chen, Payam Dowlatyari, Craig Hilby, Sazeda Sultana, Elizabeth V. Eikey, and Yunan Chen. Mapping and Taking Stock of the Personal...

    2020

  26. [34]

    Physiollm: Supporting personalized health insights with wearables and large language models

    Cathy Mengying Fang, Valdemar Danry, Nathan Whitmore, Andria Bao, Andrew Hutchison, Cayden Pierce, and Pattie Maes. Physiollm: Supporting personalized health insights with wearables and large language models. arXiv preprint arXiv:2406.19283, 2024

    2024 arXiv

  27. [35]

    Mobile sensing for behavioral research: A component-based approach for rapid deployment of sensing campaigns

    Ivan R Felix, Luis A Castro, Luis-Felipe Rodriguez, and Oresti Banos. Mobile sensing for behavioral research: A component-based approach for rapid deployment of sensing campaigns. International Journal of Distributed Sensor Networks , 15(9):1550147719874186,

  28. [36]

    Blue versus gray: A metaphor constraining sensemaking around a restructuring

    Danna N Greenberg. Blue versus gray: A metaphor constraining sensemaking around a restructuring. Group & Organization Management, 20(2):183–209, 1995. Publisher: Sage Publications Sage CA: Thousand Oaks, CA

    1995

  29. [37]

    Fostering engagement with personal informatics systems

    Rebecca Gulotta, Jodi Forlizzi, Rayoung Yang, and Mark Wah Newman. Fostering engagement with personal informatics systems. In Proceedings of the 2016 ACM conference on designing interactive systems , pages 286–300, 2016

    2016

  30. [38]

    Describing the elephant: a foundational model of human needs, motivation, behaviour, and wellbeing, December 2020

    Andreas Habermacher, Argang Ghadiri, and Theo Peters. Describing the elephant: a foundational model of human needs, motivation, behaviour, and wellbeing, December 2020

    2020

  31. [39]

    Systematic Evaluation of Personalized Deep Learning Models for Affect Recognition

    Yunjo Han, Panyu Zhang, Minseo Park, and Uichin Lee. Systematic Evaluation of Personalized Deep Learning Models for Affect Recognition. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies , 8(4):1–35, November 2024

    2024

  32. [40]

    Design of digital workplace stress-reduction intervention systems: Effects of intervention type and timing

    Esther Howe, Jina Suh, Mehrab Bin Morshed, Daniel McDuff, Kael Rowan, Javier Hernandez, Marah Ihab Abdin, Gonzalo Ramos, Tracy Tran, and Mary P Czerwinski. Design of digital workplace stress-reduction intervention systems: Effects of intervention type and timing. In Proceeding...

    2022

  33. [41]

    Large language models in mental health care: a scoping review

    Yining Hua, Fenglin Liu, Kailai Yang, Zehan Li, Hongbin Na, Yi-han Sheu, Peilin Zhou, Lauren V Moran, Sophia Ananiadou, Andrew Beam, and others. Large language models in mental health care: a scoping review. arXiv preprint arXiv:2401.02984, 2024

    2024 arXiv

  34. [42]

    Insufficient effort responding: examining an insidious confound in survey data

    Jason L Huang, Mengqiao Liu, and Nathan A Bowling. Insufficient effort responding: examining an insidious confound in survey data. Journal of Applied Psychology, 100(3):828, 2015. Publisher: American Psychological Association

    2015

  35. [43]

    Jacobson, Damien Lekkas, Raphael Huang, and Natalie Thomas

    Nicholas C. Jacobson, Damien Lekkas, Raphael Huang, and Natalie Thomas. Deep learning paired with wearable passive sensing data predicts deterioration in anxiety disorder symptoms across 17–18 years. Journal of Affective Disorders , 282:104–111, March 2021

    2021

  36. [44]

    Multimodal Autoencoder: A Deep Learning Approach to Filling In Missing Sensor Data and Enabling Better Mood Prediction

    Natasha Jaques, Sara Taylor, Akane Sano, and Rosalind Picard. Multimodal Autoencoder: A Deep Learning Approach to Filling In Missing Sensor Data and Enabling Better Mood Prediction

  37. [45]

    HARGPT: Are LLMs Zero-Shot Human Activity Recognizers? pages 38–43

    Sijie Ji, Xinzhe Zheng, and Chenshu Wu. HARGPT: Are LLMs Zero-Shot Human Activity Recognizers? pages 38–43. IEEE Computer Society, May 2024

    2024

  38. [46]

    DT2VIS: A Focus+Context Answer Generation System to Facilitate Visual Exploration of Tabular Data

    Qi Jiang, Guodao Sun, Yue Dong, and Ronghua Liang. DT2VIS: A Focus+Context Answer Generation System to Facilitate Visual Exploration of Tabular Data

  39. [47]

    Designing for data sensemaking practices: a complex challenge

    Armağan Karahanoğlu and Aykut Coşkun. Designing for data sensemaking practices: a complex challenge. Interactions, 31(4):28–31,

  40. [48]

    Drawbacks of artificial intelligence and their potential solutions in the healthcare sector.Biomedical Materials & Devices, 1(2):731–738, 2023

    Bangul Khan, Hajira Fatima, Ayatullah Qureshi, Sanjay Kumar, Abdul Hanan, Jawad Hussain, and Saad Abdullah. Drawbacks of artificial intelligence and their potential solutions in the healthcare sector.Biomedical Materials & Devices, 1(2):731–738, 2023. Publisher: Springer

    2023

  41. [49]

    Health-LLM: Large Language Models for Health Prediction via Wearable Sensor Data, April 2024

    Yubin Kim, Xuhai Xu, Daniel McDuff, Cynthia Breazeal, and Hae Won Park. Health-LLM: Large Language Models for Health Prediction via Wearable Sensor Data, April 2024. arXiv:2401.06866 [cs]

    2024 arXiv

  42. [50]

    A data–frame theory of sensemaking

    Gary Klein, Jennifer K Phillips, Erica L Rall, and Deborah A Peluso. A data–frame theory of sensemaking. In Expertise out of context , pages 118–160. Psychology Press, 2007

    2007

  43. [51]

    Intelligence essentials for everyone

    Lisa Krizan. Intelligence essentials for everyone . Joint Military Intelligence College, 1999. Number: 6

    1999

  44. [52]

    Properties and Challenges of LLM-Generated Explanations, February 2024

    Jenny Kunz and Marco Kuhlmann. Properties and Challenges of LLM-Generated Explanations, February 2024. arXiv:2402.10532 [cs]

    2024 arXiv

  45. [53]

    Why we use and abandon smart devices

    Amanda Lazar, Christian Koehler, Theresa Jean Tanenbaum, and David H Nguyen. Why we use and abandon smart devices. In Proceedings of the 2015 ACM international joint conference on pervasive and ubiquitous computing , pages 635–646, 2015

    2015

  46. [54]

    Collecting self-reported physical activity and posture data using audio-based ecological momentary assessment

    Ha Le, Rithika Lakshminarayanan, Jixin Li, Varun Mishra, and Stephen Intille. Collecting self-reported physical activity and posture data using audio-based ecological momentary assessment. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. , 8(3), September 2024. Number of ...

    2024

  47. [55]

    Feasibility and Utility of Multimodal Micro Ecological Momentary Assessment on a Smartwatch

    Ha Le, Veronika Potter, Rithika Lakshminarayanan, Varun Mishra, and Stephen Intille. Feasibility and Utility of Multimodal Micro Ecological Momentary Assessment on a Smartwatch. CHI Conference on Human Factors in Computing Systems (CHI ’25) , 2025

    2025

  48. [56]

    Rossi, and Xiang Chen

    Younghun Lee, Sungchul Kim, Tong Yu, Ryan A. Rossi, and Xiang Chen. Learning to Reduce: Optimal Representations of Structured Data in Prompting Large Language Models, February 2024. arXiv:2402.14195 [cs]

    2024 arXiv

  49. [57]

    Retrieval-augmented generation for knowledge-intensive NLP tasks

    Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, and Douwe Kiela. Retrieval-augmented generation for knowledge-intensive NLP tasks. In Proceedings of the...

    2020

  50. [58]

    Solving Quantitative Reasoning Problems with Language Models

    Aitor Lewkowycz, Anders Andreassen, David Dohan, Ethan Dyer, Henryk Michalewski, Vinay Ramasesh, Ambrose Slone, Cem Anil, Imanol Schlag, Theo Gutman-Solo, Yuhuai Wu, Behnam Neyshabur, Guy Gur-Ari, and Vedant Misra. Solving Quantitative Reasoning Problems with Language Models. ...

    2022

  51. [59]

    A stage-based model of personal informatics systems

    Ian Li, Anind Dey, and Jodi Forlizzi. A stage-based model of personal informatics systems. In Proceedings of the SIGCHI conference on human factors in computing systems , pages 557–566, 2010

    2010

  52. [60]

    Understanding my data, myself: supporting self-reflection with ubicomp technologies

    Ian Li, Anind K Dey, and Jodi Forlizzi. Understanding my data, myself: supporting self-reflection with ubicomp technologies. In Proceedings of the 13th international conference on Ubiquitous computing , pages 405–414, 2011

    2011

  53. [61]

    Vital insight: Assisting experts’ sensemaking process of multi-modal personal tracking data using visualization and LLM

    Jiachen Li, Justin Steinberg, Xiwen Li, Akshat Choube, Bingsheng Yao, Dakuo Wang, Elizabeth Mynatt, and Varun Mishra. Vital insight: Assisting experts’ sensemaking process of multi-modal personal tracking data using visualization and LLM. arXiv preprint arXiv:2410.14879, 2024

  54. [62]

    Starcoder: may the source be with you! arXiv preprint arXiv:2305.06161, 2023

    Raymond Li, Loubna Ben Allal, Yangtian Zi, Niklas Muennighoff, Denis Kocetkov, Chenghao Mou, Marc Marone, Christopher Akiki, Jia Li, Jenny Chim, and others. Starcoder: may the source be with you! arXiv preprint arXiv:2305.06161, 2023

    2023 arXiv

  55. [63]

    Just-in-time but not too much: Determining treatment timing in mobile health.Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies, 2(4):1–21, 2018

    Peng Liao, Walter Dempsey, Hillol Sarker, Syed Monowar Hossain, Mustafa Al’Absi, Predrag Klasnja, and Susan Murphy. Just-in-time but not too much: Determining treatment timing in mobile health.Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies,...

    2018

  56. [64]

    Starcoder 2 and the stack v2: The next generation

    Anton Lozhkov, Raymond Li, Loubna Ben Allal, Federico Cassano, Joel Lamy-Poirier, Nouamane Tazi, Ao Tang, Dmytro Pykhtar, Jiawei Liu, Yuxiang Wei, and others. Starcoder 2 and the stack v2: The next generation. arXiv preprint arXiv:2402.19173, 2024

    2024 arXiv

  57. [65]

    Joint Modeling of Heterogeneous Sensing Data for Depression Assessment via Multi-task Learning

    Jin Lu, Chao Shang, Chaoqun Yue, Reynaldo Morillo, Shweta Ware, Jayesh Kamath, Athanasios Bamis, Alexander Russell, Bing Wang, and Jinbo Bi. Joint Modeling of Heterogeneous Sensing Data for Depression Assessment via Multi-task Learning. Proceedings of the ACM on Interactive, M...

    2018

  58. [66]

    Breathing as Input Modality in a Gameful Breathing Training App: Development and Evaluation of Breeze 2 (Preprint)

    Yanick Xavier Lukic, Gisbert Wilhelm Teepe, Elgar Fleisch, and Tobias Kowatsch. Breathing as Input Modality in a Gameful Breathing Training App: Development and Evaluation of Breeze 2 (Preprint). preprint, JMIR Serious Games, May 2022

    2022

  59. [67]

    Accessible visualization via natural language descriptions: A four-level model of semantic content

    Alan Lundgard and Arvind Satyanarayan. Accessible visualization via natural language descriptions: A four-level model of semantic content. IEEE Transactions on Visualization and Computer Graphics , 28(1):1073–1083, January 2022

    2022

  60. [68]

    Language models of code are few-shot commonsense learners

    Aman Madaan, Shuyan Zhou, Uri Alon, Yiming Yang, and Graham Neubig. Language models of code are few-shot commonsense learners. arXiv preprint arXiv:2210.07128, 2022

    2022 arXiv

  61. [69]

    Personal discovery in diabetes self-management: discovering cause and effect using self-monitoring data

    Lena Mamykina, Elizabeth M Heitkemper, Arlene M Smaldone, Rita Kukafka, Heather J Cole-Lewis, Patricia G Davidson, Elizabeth D Mynatt, Andrea Cassells, Jonathan N Tobin, and George Hripcsak. Personal discovery in diabetes self-management: discovering cause and effect using sel...

    2017

  62. [70]

    Transforming wearable data into health insights using large language model agents.arXiv preprint arXiv:2406.06464, 2024

    Mike A Merrill, Akshay Paruchuri, Naghmeh Rezaei, Geza Kovacs, Javier Perez, Yun Liu, Erik Schenck, Nova Hammerquist, Jake Sunshine, Shyam Tailor, and others. Transforming wearable data into health insights using large language model agents.arXiv preprint arXiv:2406.06464, 2024

    2024 arXiv

  63. [71]

    Walter, Marion J

    Varun Mishra, Tian Hao, Si Sun, Kimberly N. Walter, Marion J. Ball, Ching-Hua Chen, and Xinxin Zhu. Investigating the Role of Context in Perceived Stress Detection in the Wild. In Proceedings of the 2018 ACM International Joint Conference and 2018 International Symposium on Pe...

    2018

  64. [72]

    Detecting receptivity for mHealth interventions in the natural environment

    Varun Mishra, Florian Künzler, Jan-Niklas Kramer, Elgar Fleisch, Tobias Kowatsch, and David Kotz. Detecting receptivity for mHealth interventions in the natural environment. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies , 5(2):1–24,

  65. [73]

    Continuous Detection of Physiological Stress with Commodity Hardware

    Varun Mishra, Gunnar Pope, Sarah Lord, Stephanie Lewia, Byron Lowens, Kelly Caine, Sougata Sen, Ryan Halter, and David Kotz. Continuous Detection of Physiological Stress with Commodity Hardware. ACM Transactions on Computing for Healthcare (HEALTH) , 1(2):1–30, April 2020. Pub...

    2020

  66. [74]

    Evaluating the reproducibility of physiological stress detection models

    Varun Mishra, Sougata Sen, Grace Chen, Tian Hao, Jeffrey Rogers, Ching-Hua Chen, and David Kotz. Evaluating the reproducibility of physiological stress detection models. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies , 4(4):1–29,

  67. [75]

    Personal sensing: understanding mental health using ubiquitous sensors and machine learning

    David C Mohr, Mi Zhang, and Stephen M Schueller. Personal sensing: understanding mental health using ubiquitous sensors and machine learning. Annual review of clinical psychology, 13:23–47, 2017. Publisher: Annual Reviews. , Vol. 1, No. 1, Article . Publication date: September...

    2017

  68. [76]

    Momentary stressor logging and reflective visualizations: Implications for stress management with wearables

    Sameer Neupane, Mithun Saha, Nasir Ali, Timothy Hnat, Shahin Alan Samiei, Anandatirtha Nandugudi, David M Almeida, and Santosh Kumar. Momentary stressor logging and reflective visualizations: Implications for stress management with wearables. In Proceedings of the CHI conferen...

    2024

  69. [77]

    Supporting meaningful personal fitness: The tracker goal evolution model

    Jasmin Niess and Paweł W Woźniak. Supporting meaningful personal fitness: The tracker goal evolution model. In Proceedings of the 2018 CHI conference on human factors in computing systems , pages 1–12, 2018

    2018

  70. [78]

    On the role of active memory processes in perception and cognition

    Donald A Norman and Daniel Gureasko Bobrow. On the role of active memory processes in perception and cognition . Center for Human Information Processing, Department of Psychology . . . , 1975

    1975

  71. [79]

    OpenAI documentation: Prompt engineering

    OpenAI. OpenAI documentation: Prompt engineering

  72. [80]

    GPT-4o System Card, October 2024

    OpenAI. GPT-4o System Card, October 2024. arXiv:2410.21276 [cs]

    2024 arXiv

  73. [81]

    LLMSense: Harnessing LLMs for High-level Reasoning Over Spatiotemporal Sensor Traces, March 2024

    Xiaomin Ouyang and Mani Srivastava. LLMSense: Harnessing LLMs for High-level Reasoning Over Spatiotemporal Sensor Traces, March 2024. arXiv:2403.19857 [cs]

    2024 arXiv

  74. [82]

    The sensemaking process and leverage points for analyst technology as identified through cognitive task analysis

    Peter Pirolli and Stuart Card. The sensemaking process and leverage points for analyst technology as identified through cognitive task analysis. In Proceedings of international conference on intelligence analysis , volume 5, pages 2–4. McLean, VA, USA, 2005

    2005

  75. [83]

    Behavioral indicators on a mobile sensing platform predict clinically validated psychiatric symptoms of mood and anxiety disorders

    Skyler Place, Danielle Blanch-Hartigan, Channah Rubin, Cristina Gorrostieta, Caroline Mead, John Kane, Brian P Marx, Joshua Feast, Thilo Deckersbach, Alex “Sandy” Pentland, and others. Behavioral indicators on a mobile sensing platform predict clinically validated psychiatric ...

    2017

  76. [84]

    Clear, and Peter Wright

    Aare Puussaar, Adrian K. Clear, and Peter Wright. Enhancing Personal Informatics Through Social Sensemaking. In Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems , CHI ’17, pages 6936–6942, New York, NY, USA, May 2017. Association for Computing Machinery

    2017

  77. [85]

    Building shared understanding in collaborative sensemaking

    Yan Qu and Derek L Hansen. Building shared understanding in collaborative sensemaking. In Proceedings of CHI 2008 sensemaking workshop, 2008

    2008

  78. [86]

    Lee, Ashley Garrity, and Mark W

    Shriti Raj, Joyce M. Lee, Ashley Garrity, and Mark W. Newman. Clinical data in context: Towards sensemaking tools for interpreting personal health data. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., 3(1):1–20, 2019. Publisher: ACM New York, NY, USA

    2019

  79. [87]

    CataractBot: An LLM-powered expert-in-the-loop chatbot for cataract patients

    Pragnya Ramjee, Bhuvan Sachdeva, Satvik Golechha, Shreyas Kulkarni, Geeta Fulari, Kaushik Murali, and Mohit Jain. CataractBot: An LLM-powered expert-in-the-loop chatbot for cataract patients. arXiv preprint arXiv:2402.04620, 2024

    2024 arXiv

  80. [88]

    Personal tracking as lived informatics

    John Rooksby, Mattias Rost, Alistair Morrison, and Matthew Chalmers. Personal tracking as lived informatics. In Proceedings of the SIGCHI conference on human factors in computing systems , pages 1163–1172, 2014

    2014

  81. [89]

    The cost structure of sensemaking

    Daniel M Russell, Mark J Stefik, Peter Pirolli, and Stuart K Card. The cost structure of sensemaking. In Proceedings of the INTERACT’93 and CHI’93 conference on Human factors in computing systems , pages 269–276, 1993

    1993

  82. [90]

    Stress recognition using wearable sensors and mobile phones

    Akane Sano and Rosalind W Picard. Stress recognition using wearable sensors and mobile phones. In 2013 Humaine association conference on affective computing and intelligent interaction , pages 671–676. IEEE, 2013

    2013

  83. [91]

    Introducing WESAD, a multimodal dataset for wearable stress and affect detection

    Philip Schmidt, Attila Reiss, Robert Dürichen, Claus Marberger, and Kristof Van Laerhoven. Introducing WESAD, a multimodal dataset for wearable stress and affect detection

  84. [92]

    Ecological momentary assessment

    Saul Shiffman, Arthur A Stone, and Michael R Hufford. Ecological momentary assessment. 4(1):1–32, 2008. Publisher: Annual Reviews

    2008

  85. [93]

    Towards expert-level medical question answering with large language models

    Karan Singhal, Tao Tu, Juraj Gottweis, Rory Sayres, Ellery Wulczyn, Le Hou, Kevin Clark, Stephen Pfohl, Heather Cole-Lewis, Darlene Neal, and others. Towards expert-level medical question answering with large language models. arXiv preprint arXiv:2305.09617, 2023

    2023 arXiv

  86. [94]

    Narrating fitness: Leveraging large language models for reflective fitness tracker data interpretation

    Konstantin R Strömel, Stanislas Henry, Tim Johansson, Jasmin Niess, and Paweł W Woźniak. Narrating fitness: Leveraging large language models for reflective fitness tracker data interpretation. In Proceedings of the CHI conference on human factors in computing systems, pages 1–16, 2024

    2024

  87. [95]

    Table Meets LLM: Can Large Language Models Understand Structured Table Data? A Benchmark and Empirical Study

    Yuan Sui, Mengyu Zhou, Mingjie Zhou, Shi Han, and Dongmei Zhang. Table Meets LLM: Can Large Language Models Understand Structured Table Data? A Benchmark and Empirical Study. In Proceedings of the 17th ACM International Conference on Web Search and Data Mining, WSDM ’24, pages...

    2024

  88. [96]

    DepreST-CAT: Retrospective smartphone call and text logs collected during the covid-19 pandemic to screen for mental illnesses

    ML Tlachac, Ricardo Flores, Miranda Reisch, Katie Houskeeper, and Elke A Rundensteiner. DepreST-CAT: Retrospective smartphone call and text logs collected during the covid-19 pandemic to screen for mental illnesses. Proceedings of the ACM on Interactive, Mobile, Wearable and U...

    2022

  89. [97]

    Language Models Don’t Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting

    Miles Turpin, Julian Michael, Ethan Perez, and Samuel Bowman. Language Models Don’t Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting. Advances in Neural Information Processing Systems , 36:74952–74965, December 2023

    2023

  90. [98]

    Design factors of longitudinal smartphone-based health surveys

    Sudip Vhaduri and Christian Poellabauer. Design factors of longitudinal smartphone-based health surveys. Journal of Healthcare Informatics Research, 1:52–91, 2017. Publisher: Springer

    2017

  91. [99]

    StudentLife: Assessing mental health, academic performance and behavioral trends of college students using smartphones

    Rui Wang, Fanglin Chen, Zhenyu Chen, Tianxing Li, Gabriella Harari, Stefanie Tignor, Xia Zhou, Dror Ben-Zeev, and Andrew T Campbell. StudentLife: Assessing mental health, academic performance and behavioral trends of college students using smartphones. In Proceedings of the 20...

    2014

  92. [100]

    Towards Natural Language-Based Visualization Authoring

    Yun Wang, Zhitao Hou, Leixian Shen, Tongshuang Wu, Jiaqi Wang, He Huang, Haidong Zhang, and Dongmei Zhang. Towards Natural Language-Based Visualization Authoring. IEEE Transactions on Visualization and Computer Graphics , pages 1–11, 2022. , Vol. 1, No. 1, Article . Publicatio...

    2022

  93. [101]

    MindShift: Leveraging large language models for mental-states-based problematic smartphone use intervention

    Ruolan Wu, Chun Yu, Xiaole Pan, Yujia Liu, Ningning Zhang, Yue Fu, Yuhan Wang, Zhi Zheng, Li Chen, Qiaolei Jiang, and others. MindShift: Leveraging large language models for mental-states-based problematic smartphone use intervention. In Proceedings of the CHI conference on hu...

    2024

  94. [102]

    Natural Language based Context Modeling and Reasoning for Ubiquitous Computing with Large Language Models: A Tutorial, December 2023

    Haoyi Xiong, Jiang Bian, Sijia Yang, Xiaofei Zhang, Linghe Kong, and Daqing Zhang. Natural Language based Context Modeling and Reasoning for Ubiquitous Computing with Large Language Models: A Tutorial, December 2023. arXiv:2309.15074 [cs]

    2023 arXiv

  95. [103]

    Dutcher, Yasaman S

    Xuhai Xu, Prerna Chikersal, Janine M. Dutcher, Yasaman S. Sefidgar, Woosuk Seo, Michael J. Tumminia, Daniella K. Villalba, Sheldon Cohen, Kasey G. Creswell, J. David Creswell, Afsaneh Doryab, Paula S. Nurius, Eve Riskin, Anind K. Dey, and Jennifer Mankoff. Leveraging Collabora...

    2021

  96. [104]

    GLOBEM: Cross-dataset generalization of longitudinal human behavior modeling

    Xuhai Xu, Xin Liu, Han Zhang, Weichen Wang, Subigya Nepal, Yasaman Sefidgar, Woosuk Seo, Kevin S Kuehn, Jeremy F Huckins, Margaret E Morris, and others. GLOBEM: Cross-dataset generalization of longitudinal human behavior modeling. Proceedings of the ACM on Interactive, Mobile,...

    2023

  97. [105]

    Understanding practices and needs of researchers in human state modeling by passive mobile sensing

    Xuhai Xu, Jennifer Mankoff, and Anind K Dey. Understanding practices and needs of researchers in human state modeling by passive mobile sensing. CCF Transactions on Pervasive Computing and Interaction , 3:344–366, 2021. Publisher: Springer

    2021

  98. [106]

    GLOBEM dataset: Multi-year datasets for longitudinal human behavior modeling generalization

    Xuhai Xu, Han Zhang, Yasaman Sefidgar, Yiyi Ren, Xin Liu, Woosuk Seo, Jennifer Brown, Kevin Kuehn, Mike Merrill, Paula Nurius, and others. GLOBEM dataset: Multi-year datasets for longitudinal human behavior modeling generalization. Advances in Neural Information Processing Sys...

    2022

  99. [107]

    DrHouse: An LLM-empowered Diagnostic Reasoning System through Harnessing Outcomes from Sensor Data and Expert Knowledge

    Bufang Yang, Siyang Jiang, Lilin Xu, Kaiwei Liu, Hai Li, Guoliang Xing, Hongkai Chen, Xiaofan Jiang, and Zhenyu Yan. DrHouse: An LLM-empowered Diagnostic Reasoning System through Harnessing Outcomes from Sensor Data and Expert Knowledge. Proc. ACM Interact. Mob. Wearable Ubiqu...

    2024

  100. [108]

    Bitterman, Jasmine Chiat Ling Ong, Daniel Shu Wei Ting, and Nan Liu

    Rui Yang, Yilin Ning, Emilia Keppo, Mingxuan Liu, Chuan Hong, Danielle S. Bitterman, Jasmine Chiat Ling Ong, Daniel Shu Wei Ting, and Nan Liu. Retrieval-augmented generation for generative artificial intelligence in health care. npj Health Systems, 2(1):1–5, January

  101. [109]

    Why Johnny can’t prompt: How non-AI experts try (and fail) to design LLM prompts

    JD Zamfirescu-Pereira, Richmond Y Wong, Bjoern Hartmann, and Qian Yang. Why Johnny can’t prompt: How non-AI experts try (and fail) to design LLM prompts. In Proceedings of the 2023 CHI conference on human factors in computing systems , pages 1–21, 2023

    2023

  102. [110]

    Towards a comprehensive model of the cognitive process and mechanisms of individual sensemaking

    Pengyi Zhang and Dagobert Soergel. Towards a comprehensive model of the cognitive process and mechanisms of individual sensemaking. Journal of the Association for Information Science and Technology , 65(9):1733–1756, 2014. Publisher: Wiley Online Library

    2014

  103. [111]

    A survey on neural network interpretability

    Yu Zhang, Peter Tiňo, Aleš Leonardis, and Ke Tang. A survey on neural network interpretability. IEEE Transactions on Emerging Topics in Computational Intelligence, 5(5):726–742, 2021. Publisher: IEEE. , Vol. 1, No. 1, Article . Publication date: September 2025

    2021

  104. [2019]

    Publisher: SAGE Publications Sage UK: London, England

  105. [2020]

    Publisher: ACM New York, NY, USA

  106. [2024]

    Publisher: ACM New York, NY, USA tex.issue_date: July - August 2024

    2024

  107. [2025]

    Publisher: Nature Publishing Group

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.