arXiv subjects

John Mitchell

Publications and source records attributed to John Mitchell.

At least 19 recordsLinked to original sources

arXiv2026

How AI Coders Discuss, Disagree, and Reach Consensus: Challenges and Opportunities for LLM-Based Qualitative Coding

The utility of AI in multi-coder qualitative coding has been widely discussed, yet little empirical evidence exists to delineate the contexts in which it performs reliably. We address this gap by quantifying the effectiveness of multi-agent LLM coding across varied qualitative datasets, revealing key contextual and structural factors that mediate coding outcomes. We developed a literature-informed baseline pipeline that enables AI agents to independently code, debate, and reconcile disagreements. Results revealed that coding accuracy depends on factors such as codebook length, qualitative data similarity, and agent disagreement. Notably, intense and unresolved debates between agents led to higher accuracy. Our analysis showed that while LLMs emulate many human discussion behaviors, they lack adaptive responsiveness to context. From these findings, we offer design recommendations for building automated coding systems. Our open-source AI discussion dataset and methodological framework lay the groundwork for advancing the design of AI-mediated automated thematic analysis.

cs.HC↗

arXiv2026

Integrating Virtual Reality and Large Language Models for Team-Based Non-Technical Skills Training and Evaluation in the Operating Room

Although effective teamwork and communication are critical to surgical safety, structured training for non-technical skills (NTS) remains limited compared with technical simulation. The ACS/APDS Phase III Team-Based Skills Curriculum calls for scalable tools that both teach and objectively assess these competencies during laparoscopic emergencies. We introduce the Virtual Operating Room Team Experience (VORTeX), a multi-user virtual reality (VR) platform that integrates immersive team simulation with large language model (LLM) analytics to train and evaluate communication, decision-making, teamwork, and leadership. Team dialogue is analyzed using structured prompts derived from the Non-Technical Skills for Surgeons (NOTSS) framework, enabling automated classification of behaviors and generation of directed interaction graphs that quantify communication structure and hierarchy. Two laparoscopic emergency scenarios, pneumothorax and intra-abdominal bleeding, were implemented to elicit realistic stress and collaboration. Twelve surgical professionals completed pilot sessions at the 2024 SAGES conference, rating VORTeX as intuitive, immersive, and valuable for developing teamwork and communication. The LLM consistently produced interpretable communication networks reflecting expected operative hierarchies, with surgeons as central integrators, nurses as initiators, and anesthesiologists as balanced intermediaries. By integrating immersive VR with LLM-driven behavioral analytics, VORTeX provides a scalable, privacy-compliant framework for objective assessment and automated, data-informed debriefing across distributed training environments.

cs.HC↗

arXiv2025

Structural origin of resonant diffraction in RuO$_2$

We report Ru L$_3$-edge resonant X-ray diffraction studies on single crystal and (001) epitaxial films of RuO$_2$. We investigate the distinct $\mathbf{Q} = (100)$ and $(001)$ reflections as a function of incident energy, azimuthal angle, and temperature. The results show that the observed resonant diffraction in RuO$_2$ is fully consistent with a resonant charge anisotropy signal of structural origin permitted by the parent (non-magnetic) rutile $P4_2/mnm$ space group. These results significantly constrain the magnetic contribution to the resonant diffraction signal and indicate the unlikely existence of $\mathbf{k} = 0$ antiferromagnetic order in RuO$_2$.

cond-mat.str-el↗

arXiv2025

Calibration of the ComPair Balloon Instrument

ComPair, the prototype of the All-sky Medium Energy Gamma-ray Observatory (AMEGO) mission concept, is a combined Compton imager and pair production telescope. It consists of four subsystems: a double-sided silicon strip detector (DSSD) Tracker, a virtual Frisch-grid cadmium zinc telluride (CZT) Low Energy Calorimeter, a cesium iodide (CsI) High Energy Calorimeter, and a plastic scintillator Anti-Coincidence Detector (ACD) to reject the charged particle background. These subsystems work together to reconstruct events, by tracking the locations and energies of gamma-ray scatters and pair production events. To quantify ComPair's scientific capabilities prior to a balloon launch in 2023, calibrations were performed to benchmark the instrument's performance in terms of angular resolution, energy resolution, and effective area. In this paper we provide an overview of the ComPair instrument and detail the calibration campaign. Finally, we compare our results to the expected performance based on simulations.

astro-ph.IM↗

arXiv2025

Background Measurements and Simulations of the ComPair Balloon Flight

ComPair, a prototype of the All-sky Medium Energy Gamma-ray Observatory (AMEGO), completed a short-duration high-altitude balloon campaign on August 27, 2023 from Fort Sumner, New Mexico, USA. The goal of the balloon flight was the demonstration of ComPair as both a Compton and Pair telescope in flight, rejection of the charged particle background, and measurement of the background $\gamma$-ray spectrum. This analysis compares measurements from the balloon flight with Monte Carlo simulations to benchmark the instrument. The comparison finds good agreement between the measurements and simulations and supports the conclusion that ComPair accomplished its goals for the balloon campaign. Additionally, two charged particle background rejection schemes are discussed: a soft ACD veto that records a higher charged particle event rate but with less risk of event loss, and a hard ACD veto that limits the charged particle event rate on board. There was little difference in the measured spectra from the soft and hard ACD veto schemes, indicating that the hard ACD veto could be used for future flights. The successes of ComPair's engineering flight will inform the development of the next generation of ComPair with upgraded detector technology and larger active area.

astro-ph.IM↗

arXiv2024

ComPair-2: A Next Generation Medium Energy Gamma-ray Telescope Prototype

Many questions posed in the Astro2020 Decadal survey in both the New Messengers and New Physics and the Cosmic Ecosystems science themes require a gamma-ray mission with capabilities exceeding those of existing (e.g. Fermi, Swift) and planned (e.g. COSI) observatories. ComPair, the Compton Pair telescope, is a prototype of such a next-generation gamma-ray mission. It had its inaugural balloon flight from Ft. Sumner, New Mexico in August 2023. To continue the goals of the ComPair project to develop technologies that will enable a future gamma-ray mission, the next generation of ComPair (ComPair-2) will be upgraded to increase the sensitivity and low-energy transient capabilities of the instrument. These advancements are enabled by AstroPix, a silicon monolithic active pixel sensor, in the tracker and custom dual-gain silicon photomultipliers and front-end electronics in the calorimeter. This effort builds on design work for the All-sky Medium Energy Gamma-ray Observatory eXplorer (AMEGO-X) concept that was submitted the 2021 MIDEX Announcement of Opportunity. Here we describe the ComPair-2 prototype design and integration and testing plans to advance the readiness level of these novel technologies.

astro-ph.IM↗

arXiv2024

Origin of the Unusual Temperature Dependence of the Upper Critical Field of Kagome Superconductor CsV3Sb5: Multiple Bands or van Hove Singularities?

Van Hove singularities (vHs) located close to the Fermi level in Kagome superconductors AV3Sb5 (A = K, Rb, Cs) have profound influence on their electronic and transport characteristics. Specifically, magneto-transport and susceptibility measurements on CsV3Sb5 reveal an anomalous temperature dependence of the upper critical field H_c2 (T), characterized by a pronounced upward curvature for both in-plane and c-axis magnetic fields, with zero-temperature H_c2 values of ~6.0 T and ~1.2 T, respectively. Our theoretical analysis, using a newly developed single-band model incorporating vHs and gap anisotropy, suggests that the observed upper critical field behavior is predominantly driven by the anisotropy of the Fermi velocity originating from vHs, instead of multi-band effects or gap anisotropy. Increased electron scattering introduced by proton irradiation defects smears out the vHs, reduces anisotropy, and recovers the conventional H_c2 (T) behavior, corroborating our proposed model.

cond-mat.supr-con↗

arXiv2024

The 2023 Balloon Flight of the ComPair Instrument

The ComPair balloon instrument is a prototype gamma-ray telescope that aims to further develop technology for observing the gamma-ray sky in the MeV regime. ComPair combines four detector subsystems to enable parallel Compton scattering and pair-production detection, critical for observing in this energy range. This includes a 10 layer double-sided silicon strip detector tracker, a virtual Frisch grid low energy CZT calorimeter, a high energy CsI calorimeter, and a plastic scintillator anti-coincidence detector. The inaugural balloon flight successfully launched from the Columbia Scientific Balloon Facility site in Fort Sumner, New Mexico, in late August 2023, lasting approximately 6.5 hours in duration. In this proceeding, we discuss the development of the ComPair Since balloon payload, the performance during flight, and early results.

astro-ph.IM↗

arXiv2024

Securing the Future of GenAI: Policy and Technology

The rise of Generative AI (GenAI) brings about transformative potential across sectors, but its dual-use nature also amplifies risks. Governments globally are grappling with the challenge of regulating GenAI, balancing innovation against safety. China, the United States (US), and the European Union (EU) are at the forefront with initiatives like the Management of Algorithmic Recommendations, the Executive Order, and the AI Act, respectively. However, the rapid evolution of GenAI capabilities often outpaces the development of comprehensive safety measures, creating a gap between regulatory needs and technical advancements. A workshop co-organized by Google, University of Wisconsin, Madison (UW-Madison), and Stanford University aimed to bridge this gap between GenAI policy and technology. The diverse stakeholders of the GenAI space -- from the public and governments to academia and industry -- make any safety measures under consideration more complex, as both technical feasibility and regulatory guidance must be realized. This paper summarizes the discussions during the workshop which addressed questions, such as: How regulation can be designed without hindering technological progress? How technology can evolve to meet regulatory standards? The interplay between legislation and technology is a very vast topic, and we don't claim that this paper is a comprehensive treatment on this topic. This paper is meant to capture findings based on the workshop, and hopefully, can guide discussion on this topic.

cs.CY↗

arXiv2024

Results from the CsI Calorimeter onboard the 2023 ComPair Balloon Flight

The ComPair gamma-ray telescope is a technology demonstrator for a future gamma-ray telescope called the All-sky Medium Energy Gamma-ray Observatory (AMEGO). The instrument is composed of four subsystems, a double-sided silicon strip detector, a virtual Frisch grid CdZnTe calorimeter, a CsI:Tl based calorimeter, and an anti-coincidence detector (ACD). The CsI calorimeter's goal is to measure the position and energy deposited from high-energy events. To demonstrate the technological readiness, the calorimeter has flown onboard a NASA scientific balloon as part of the GRAPE-ComPair mission and accumulated around 3 hours of float time at an altitude of 40 km. During the flight, the CsI calorimeter observed background radiation, Regener-Pfotzer Maximum, and several gamma-ray activation lines originating from aluminum.

astro-ph.IM↗

arXiv2024

Social Skill Training with Large Language Models

People rely on social skills like conflict resolution to communicate effectively and to thrive in both work and personal life. However, practice environments for social skills are typically out of reach for most people. How can we make social skill training more available, accessible, and inviting? Drawing upon interdisciplinary research from communication and psychology, this perspective paper identifies social skill barriers to enter specialized fields. Then we present a solution that leverages large language models for social skill training via a generic framework. Our AI Partner, AI Mentor framework merges experiential learning with realistic practice and tailored feedback. This work ultimately calls for cross-disciplinary innovation to address the broader implications for workforce development and social equality.

cs.CL↗

arXiv2024

DPAdapter: Improving Differentially Private Deep Learning through Noise Tolerance Pre-training

Recent developments have underscored the critical role of \textit{differential privacy} (DP) in safeguarding individual data for training machine learning models. However, integrating DP oftentimes incurs significant model performance degradation due to the perturbation introduced into the training process, presenting a formidable challenge in the {differentially private machine learning} (DPML) field. To this end, several mitigative efforts have been proposed, typically revolving around formulating new DPML algorithms or relaxing DP definitions to harmonize with distinct contexts. In spite of these initiatives, the diminishment induced by DP on models, particularly large-scale models, remains substantial and thus, necessitates an innovative solution that adeptly circumnavigates the consequential impairment of model utility. In response, we introduce DPAdapter, a pioneering technique designed to amplify the model performance of DPML algorithms by enhancing parameter robustness. The fundamental intuition behind this strategy is that models with robust parameters are inherently more resistant to the noise introduced by DP, thereby retaining better performance despite the perturbations. DPAdapter modifies and enhances the sharpness-aware minimization (SAM) technique, utilizing a two-batch strategy to provide a more accurate perturbation estimate and an efficient gradient descent, thereby improving parameter robustness against noise. Notably, DPAdapter can act as a plug-and-play component and be combined with existing DPML algorithms to further improve their performance. Our experiments show that DPAdapter vastly enhances state-of-the-art DPML algorithms, increasing average accuracy from 72.92\% to 77.09\% with a privacy budget of $\epsilon=4$.

cs.LG↗

arXiv2024

TrustLLM: Trustworthiness in Large Language Models

Large language models (LLMs), exemplified by ChatGPT, have gained considerable attention for their excellent natural language processing capabilities. Nonetheless, these LLMs present many challenges, particularly in the realm of trustworthiness. Therefore, ensuring the trustworthiness of LLMs emerges as an important topic. This paper introduces TrustLLM, a comprehensive study of trustworthiness in LLMs, including principles for different dimensions of trustworthiness, established benchmark, evaluation, and analysis of trustworthiness for mainstream LLMs, and discussion of open challenges and future directions. Specifically, we first propose a set of principles for trustworthy LLMs that span eight different dimensions. Based on these principles, we further establish a benchmark across six dimensions including truthfulness, safety, fairness, robustness, privacy, and machine ethics. We then present a study evaluating 16 mainstream LLMs in TrustLLM, consisting of over 30 datasets. Our findings firstly show that in general trustworthiness and utility (i.e., functional effectiveness) are positively related. Secondly, our observations reveal that proprietary LLMs generally outperform most open-source counterparts in terms of trustworthiness, raising concerns about the potential risks of widely accessible open-source LLMs. However, a few open-source LLMs come very close to proprietary ones. Thirdly, it is important to note that some LLMs may be overly calibrated towards exhibiting trustworthiness, to the extent that they compromise their utility by mistakenly treating benign prompts as harmful and consequently not responding. Finally, we emphasize the importance of ensuring transparency not only in the models themselves but also in the technologies that underpin trustworthiness. Knowing the specific trustworthy technologies that have been employed is crucial for analyzing their effectiveness.

cs.CL↗

arXiv2023

Identifying and Mitigating the Security Risks of Generative AI

Every major technical invention resurfaces the dual-use dilemma -- the new technology has the potential to be used for good as well as for harm. Generative AI (GenAI) techniques, such as large language models (LLMs) and diffusion models, have shown remarkable capabilities (e.g., in-context learning, code-completion, and text-to-image generation and editing). However, GenAI can be used just as well by attackers to generate new attacks and increase the velocity and efficacy of existing attacks. This paper reports the findings of a workshop held at Google (co-organized by Stanford University and the University of Wisconsin-Madison) on the dual-use dilemma posed by GenAI. This paper is not meant to be comprehensive, but is rather an attempt to synthesize some of the interesting findings from the workshop. We discuss short-term and long-term goals for the community on this topic. We hope this paper provides both a launching point for a discussion on this important topic as well as interesting problems that the research community can work to address.

cs.AI↗

arXiv2023

The Compton-Pair telescope: A prototype for a next-generation MeV $\gamma$-ray observatory

The Compton Pair (ComPair) telescope is a prototype that aims to develop the necessary technologies for future medium energy gamma-ray missions and to design, build, and test the prototype in a gamma-ray beam and balloon flight. The ComPair team has built an instrument that consists of 4 detector subsystems: a double-sided silicon strip detector Tracker, a novel high-resolution virtual Frisch-grid cadmium zinc telluride Calorimeter, and a high-energy hodoscopic cesium iodide Calorimeter, all of which are surrounded by a plastic scintillator anti-coincidence detector. These subsystems together detect and characterize photons via Compton scattering and pair production, enable a veto of cosmic rays, and are a proof-of-concept for a space telescope with the same architecture. A future medium-energy gamma-ray mission enabled through ComPair will address many questions posed in the Astro2020 Decadal survey in both the New Messengers and New Physics and the Cosmic Ecosystems themes. In this contribution, we will give an overview of the ComPair project and steps forward to the balloon flight.

astro-ph.IM↗

arXiv2023

Symbiotic and antagonistic disease dynamics on networks using bond percolatio

In this paper we introduce a novel description of the equilibrium state of a bond percolation process on random graphs using the exact method of generating functions. This allows us to find the expected size of the giant connected component (GCC) of two sequential bond percolation processes in which the bond occupancy probability of the second process is modulated (increased or decreased) by a node being inside or outside of the GCC created by the first process. In the context of epidemic spreading this amounts to both a antagonistic partial immunity or a synergistic partial coinfection interaction between the two sequential diseases. We examine configuration model networks with tunable clustering. We find that the emergent evolutionary behaviour of the second strain is highly dependent on the details of the coupling between the strains. Contact clustering generally reduces the outbreak size of the second strain relative to unclustered topologies; however, positive assortativity induced by clustered contacts inverts this conclusion for highly transmissible disease dynamics.

physics.soc-ph↗

arXiv2022

Development of the ComPair gamma-ray telescope prototype

There is a growing interest in the science uniquely enabled by observations in the MeV range, particularly in light of multi-messenger astrophysics. The Compton Pair (ComPair) telescope, a prototype of the AMEGO Probe-class concept, consists of four subsystems that together detect and characterize gamma rays in the MeV regime. A double-sided strip silicon Tracker gives a precise measure of the first Compton scatter interaction and tracks pair-conversion products. A novel cadmium zinc telluride (CZT) detector with excellent position and energy resolution beneath the Tracker detects the Compton-scattered photons. A thick cesium iodide (CsI) calorimeter contains the high-energy Compton and pair events. The instrument is surrounded by a plastic anti-coincidence (ACD) detector to veto the cosmic-ray background. In this work, we will give an overview of the science motivation and a description of the prototype development and performance.

astro-ph.IM↗

arXiv2022

The Move Borrow Checker

The Move language provides abstractions for programming with digital assets via a mix of value semantics and reference semantics. Ensuring memory safety in programs with references that access a shared, mutable global ledger is difficult, yet essential for the use-cases targeted by Move. The language meets this challenge with a novel memory model and a modular, intraprocedural static reference safety analysis that leverages key properties of the memory. The analysis ensures the absence of memory safety violations in all Move programs (including ones that link against untrusted code) by running as part of a load-time bytecode verification pass similar to the JVM [12] and CLR [15]. We formalize the static analysis and prove that it enjoys three desirable properties: absence of dangling references, referential transparency for immutable references, and absence of memory leaks.

cs.PL↗