Department of Computer Science and Engineering Faculty Papers

Permanent URI for this collectionhttps://digital.lib.washington.edu/handle/1773/15625

Browse

Now showing 1 - 6 of 6

Position: Knowing Isn't Understanding: Re-grounding Generative Proactivity with Epistemic and Behavioral Insight
(2026-02-03) Kaur, Kirandeep; Lyu, Xingda; Shah, Chirag
Generative AI agents equate understanding with resolving explicit queries, an assumption that confines interaction to what users can articulate. This assumption breaks down when users themselves lack awareness of what is missing, risky, or worth considering. In such conditions, proactivity is not merely an efficiency enhancement, but an epistemic necessity. We refer to this condition as epistemic incompleteness, where progress depends on engaging with unknown \textit{unknowns} for effective partnership. Existing approaches to proactivity remain narrowly anticipatory, extrapolating from past behavior and presuming that goals are already well defined, thereby failing to support users meaningfully. However, surfacing possibilities beyond a user’s current awareness is not inherently beneficial. Unconstrained proactive interventions can misdirect attention, overwhelm users, or introduce harm. Proactive agents, therefore, require behavioral grounding: principled constraints on when, how, and to what extent an agent should intervene. We advance the position that generative proactivity must be grounded both epistemically and behaviorally. Drawing on the philosophy of ignorance and research on proactive behavior, we argue that these theories offer critical guidance for designing agents that can engage responsibly and foster meaningful partnerships.
Benchmarking and Advancing Knowledge Gap Navigation
(2026-01-26) Kaur, Kirandeep; Gupta , Vinayak; Gupta , Aditya; Shah, Chirag
Most language-based assistants follow a reactive ask-and-respond paradigm, requiring users to explicitly state their needs. As a result, relevant but unexpressed needs often go unmet. Existing proactive agents attempt to address this gap either by eliciting further clarification, preserving this burden, or by extrapolating future needs from context, often leading to unnecessary or mistimed interventions. We introduce PROPER, Proactivity-driven Personalized agents, a novel two-agent architecture consisting of a Dimension Generating Agent (DGA) and a Response Generating Agent (RGA). DGA, a fine-tuned LLM agent, leverages explicit user data to generate multiple implicit dimensions (latent aspects relevant to the user’s task but not considered by the user) or knowledge gaps. These dimensions are selectively filtered using a reranker based on quality, diversity, and task-relevance. RGA then balances explicit and implicit dimensions to tailor personalized responses with timely and proactive interventions. We evaluate PROPER across multiple domains using a structured, gap-aware rubric that measures coverage, initiative appropriateness, and intent alignment. Our results show that PROPER improves on quality scores and win rates across all domains, achieving up to 84% gains in single-turn evaluation and consistent dominance in multi-turn interactions.
Six Solutions to the Money Supply Problem
(2025-09-01) Rao, Anup
In the money supply problem, the goal is to design a mechanism for controlling the money supply that results in the least distortion of prices. We study 6 solutions to this problem, including 2 novel proposals based on an auction mechanism.
Decentralized Money Supply: a New Paradigm for Reserve Currency
(2025-05-20) Rao, Anup
This paper proposes a decentralized reserve currency called the Global Dollar. Global Dollars are compatible with the traditional banking system, and yet have a completely decentralized mechanism for regulating the money supply. The goal is to provide a viable alternative to existing global reserve currencies by maintaining decentralization while incorporating the best features of existing reserve currencies.
Proceedings of the Workshop on Designing Technologies to Support Human Problem Solving
(2018-10-01) Tanimoto, Steven; Fan, Sandra; Ko, Andrew; Locksa, Dastyni
Conference Proceedings from DTSHPS: Lisbon, Portugal on October 1, 2018, in conjunction with the IEEE Symposium on Visual Languages and Human Centric Computing.
Randomized clinical trial of surgery versus conservative therapy for carpal tunnel syndrome [ISRCTN84286481]
(2005) Martin, Brook I.; Levenson, Linda M.; Hollingworth, William; Kliot, Michel; Heagerty, Patrick J.; Turner, Judith A.; Jarvik, Jeffrey G.
Background: Conservative treatment remains the standard of care for treating mild to moderate carpal tunnel syndrome despite a small number of well-controlled studies and limited objective evidence to support current treatment options. There is an increasing interest in the usefulness of wrist magnetic resonance imaging could play in predicting who will benefit for various treatments. Method and design: Two hundred patients with mild to moderate symptoms will be recruited over 3 1/2 years from neurological surgery, primary care, electrodiagnostic clinics. We will exclude patients with clinical or electrodiagnostic evidence of denervation or thenar muscle atrophy. We will randomly assign patients to either a well-defined conservative care protocol or surgery. The conservative care treatment will include visits with a hand therapist, exercises, a self-care booklet, work modification/ activity restriction, B6 therapy, ultrasound and possible steroid injections. The surgical care would be left up to the surgeon (endoscopic vs. open) with usual and customary follow-up. All patients will receive a wrist MRI at baseline. Patients will be contacted at 3, 6, 9 and 12 months after randomization to complete the Carpal Tunnel Syndrome Assessment Questionnaire (CTSAQ). In addition, we will compare disability (activity and work days lost) and general well being as measured by the SF-36 version II. We will control for demographics and use psychological measures (SCL-90 somatization and depression scales) as well as EDS and MRI predictors of outcomes. Discussion: We have designed a randomized controlled trial which will assess the effectiveness of surgery for patients with mild to moderate carpal tunnel syndrome. An important secondary goal is to study the ability of MRI to predict patient outcomes.

Browse

Recent Submissions