Search

We present the results of a Keck and NOEMA spectroscopic survey of 507 galaxies, where we confirm the presence of two massive overdensities at $z = 3.090 - 3.110$ and $z = 3.133 - 3.155$ in the neighborhood of the GOODS-N, each with over a dozen spectroscopically confirmed members. We find that both of these have galaxy overdensities of NIR-detected galaxies of $\delta_{\rm gal, obs} = 6 - 9$ within corrected volumes of $(6 - 7) \times 10^3~{\rm cMpc}^3$. We estimate the properties of the $z = 0$ descendants of these overdensities using a spherical collapse model and find that both should virialize by $z \simeq 0.5 - 0.8$, with total masses of $M_{\rm tot} \simeq (6 - 7) \times 10^{14}~{\rm M}_\odot$. The same spherical collapse calculations, as well as a clustering-of-clusters statistical analysis, suggest a >80% likelihood that the two overdensities will collapse into a single cluster with $M_{\rm tot} = (1.0 - 1.5) \times 10^{15}~{\rm M}_\odot$ by $z \sim 0.1-0.4$. The $z = 3.14$ substructure contains a core of four bright dusty star-forming galaxies with $\Sigma {\rm SFR} = 2700 \pm 700~{\rm M}_\odot~{\rm yr}^{-1}$ in a volume of only 280 ${\rm cMpc}^3$., Comment: 18 pages, 8 figures. Accepted to ApJ

We present a 34-year timing solution of the redback pulsar system Terzan 5A (Ter5A). Ter5A, also known as B1744$-$24A or J1748$-$2446A, has a 11.56 ms pulse period, a $\sim$0.1 solar mass dwarf companion star, and an orbital period of 1.82 hours. Ter5A displays highly variable eclipses and orbital perturbations. Using new timing techniques, we have determined a phase-connected timing solution for this system over 34 years. This is the longest ever published for a redback pulsar. We find that the pulsar's spin variability is much larger than most globular cluster pulsars. In fact, of the nine redback pulsars with published or in preparation long-term timing solutions, Ter5A is by far the noisiest. We see no evidence of strong correlations between orbital and spin variability of the pulsar. We also find that long-term astrometric timing measurements are likely too contaminated by this variability to be usable, and therefore require careful short-term timing to determine reasonable positions. Finally, we measure an orbital period contraction of $-2.5(3) \times 10^{-13}$, which is likely dominated by the general relativistic orbital decay of the system. The effects of the orbital variability due to the redback nature of the pulsar are not needed to explain the observed orbital period derivative, but they are constrained to less than $\sim$30% of the observed value.

With plasma currents up to 8.7 MA, the SPARC tokamak runs the risk of forming multi-MA beams of relativistic "runaway" electrons (REs) which could damage plasma facing components if unmitigated. The infrared (IR) and visible imaging and visible spectroscopy systems in SPARC are designed with measurements of synchrotron emission from REs in mind. Synchrotron radiation is emitted by REs along their direction of motion, opposite the plasma current. Matched clockwise and counterclockwise wide views are proposed to detect synchrotron and background radiation, allowing observation of RE synchrotron emission in both plasma current configurations. Due to SPARC's high toroidal magnetic field strength, 12.2 T on axis, the synchrotron light spectrum is expected to peak in the visible-IR wavelength range. The synthetic diagnostic tool SOFT (Synchrotron Orbit-Following Toolkit) is used to model synchrotron images and spectra for three scenarios, with appropriate magnetic equilibria for each: REs generated during plasma current ramp-up, steady-state flattop (although unlikely, but serving as a reference), and disruptions. Required time resolutions, achievable spatial coverage, and appropriate spectral ranges for various RE energies are assessed.

We study the complexity of testing properties of quantum channels. First, we show that testing identity to any channel $\mathcal N: \mathbb C^{d_{\mathrm{in}} \times d_{\mathrm{in}}} \to \mathbb C^{d_{\mathrm{out}} \times d_{\mathrm{out}}}$ in diamond norm distance requires $\Omega(\sqrt{d_{\mathrm{in}}} / \varepsilon)$ queries, even in the strongest algorithmic model that admits ancillae, coherence, and adaptivity. This is due to the worst-case nature of the distance induced by the diamond norm. Motivated by this limitation and other theoretical and practical applications, we introduce an average-case analogue of the diamond norm, which we call the average-case imitation diamond (ACID) norm. In the weakest algorithmic model without ancillae, coherence, or adaptivity, we prove that testing identity to certain types of channels in ACID distance can be done with complexity independent of the dimensions of the channel, while for other types of channels the complexity depends on both the input and output dimensions. Building on previous work, we also show that identity to any fixed channel can be tested with $\tilde O(d_{\mathrm{in}} d_{\mathrm{out}}^{3/2} / \varepsilon^2)$ queries in ACID distance and $\tilde O(d_{\mathrm{in}}^2 d_{\mathrm{out}}^{3/2} / \varepsilon^2)$ queries in diamond distance in this model. Finally, we prove tight bounds on the complexity of channel tomography in ACID distance.

Machine Learning as a Service (MLaaS) is often provided as a pay-per-query, black-box system to clients. Such a black-box approach not only hinders open replication, validation, and interpretation of model results, but also makes it harder for white-hat researchers to identify vulnerabilities in the MLaaS systems. Model extraction is a promising technique to address these challenges by reverse-engineering black-box models. Since training data is typically unavailable for MLaaS models, this paper focuses on the realistic version of it: data-free model extraction. We propose a data-free model extraction approach, CaBaGe, to achieve higher model extraction accuracy with a small number of queries. Our innovations include (1) a novel experience replay for focusing on difficult training samples; (2) an ensemble of generators for steadily producing diverse synthetic data; and (3) a selective filtering process for querying the victim model with harder, more balanced samples. In addition, we create a more realistic setting, for the first time, where the attacker has no knowledge of the number of classes in the victim training data, and create a solution to learn the number of classes on the fly. Our evaluation shows that CaBaGe outperforms existing techniques on seven datasets -- MNIST, FMNIST, SVHN, CIFAR-10, CIFAR-100, ImageNet-subset, and Tiny ImageNet -- with an accuracy improvement of the extracted models by up to 43.13%. Furthermore, the number of queries required to extract a clone model matching the final accuracy of prior work is reduced by up to 75.7%.

For random-walk Metropolis (RWM) and parallel tempering (PT) algorithms, an asymptotic acceptance rate of around 0.234 is known to be optimal in the high-dimensional limit. However, its practical relevance is uncertain due to restrictive derivation conditions. We synthesise previous theoretical advances in extending the 0.234 acceptance rate to more general settings, and demonstrate its applicability with a comprehensive empirical simulation study on examples examining how acceptance rates affect Expected Squared Jumping Distance (ESJD). Our experiments show the optimality of the 0.234 acceptance rate for RWM is surprisingly robust even in lower dimensions across various proposal and multimodal target distributions which may or may not have an i.i.d. product density. Parallel tempering experiments also show that the idealized 0.234 spacing of inverse temperatures may be approximately optimal for low dimensions and non i.i.d. product target densities, and that constructing an inverse temperature ladder with spacings given by a swap acceptance of 0.234 is a viable strategy. However, we observe the applicability of the 0.234 acceptance rate heuristic diminishes for both RWM and PT algorithms below a certain dimension which differs based on the target density, and that inhomogeneously scaled components in the target density further reduces its applicability in lower dimensions., Comment: Submitted to Communications in Statistics - Simulation and Computation

In this bachelor's thesis, isotope effects for pedestal instabilities have been studied based on an ASDEX Upgrade H-Mode scenario. This was done using the JOREK code for extended MHD simulations, including the ion diamagnetic drift and the establishment of ExB flows. Simulations with single toroidal harmonics were performed for multiple times during the build-up of the pedestal, to assess the evolution of the linear stability of modes occurring near the edge. When changing the average ion masses from 2.0 to 2.5 and 3.0, the variations were small, and MHD's momentum equation was shown to be the major cause for them. As a second step, simulations with multiple toroidal harmonics were performed to simulate an Edge Localized Mode (ELM) crash, again comparing between the average ion masses of 2.0, 2.5, and 3.0. The resulting variations of heat and particle losses were compared to JET results. The experimental differences between ion masses could not be reproduced. Together with the fact that the simulations' pedestal for an average ion mass of 2.5 and 3.0 was not matched to the experiment, this indicates that other effects, such as small-scale turbulences, must be included - for example, by adapting the simulations' transport coefficients - to explain the experimental differences.

We review and provide new proofs of results used to compare the efficiency of estimates generated by reversible MCMC algorithms on a general state space. We provide a full proof of the formula for the asymptotic variance for real-valued functionals on $\phi$-irreducible reversible Markov chains, first introduced by Kipnis and Varadhan. Given two Markov kernels $P$ and $Q$ with stationary measure $\pi$, we say that the Markov kernel $P$ efficiency dominates the Markov kernel $Q$ if the asymptotic variance with respect to $P$ is at most the asymptotic variance with respect to $Q$ for every real-valued functional $f\in L^2(\pi)$. Assuming only a basic background in functional analysis, we prove that for two $\phi$-irreducible reversible Markov kernels $P$ and $Q$, $P$ efficiency dominates $Q$ if and only if the operator $Q-P$, where $P$ is the operator on $L^2(\pi)$ that maps $f\mapsto\int f(y)P(\cdot,dy)$ and similarly for $Q$, is positive on $L^2(\pi)$, i.e. $\langle f,(Q-P)f\rangle\geq0$ for every $f\in L^2(\pi)$. We use this result to show that reversible antithetic kernels are more efficient than i.i.d. sampling, and that efficiency dominance is a partial ordering on $\phi$-irreducible reversible Markov kernels. We also provide a proof based on that of Tierney that Peskun dominance is a sufficient condition for efficiency dominance for reversible kernels. Using these results, we show that Markov kernels formed by randomly selecting other "component" Markov kernels will always efficiency dominate another Markov kernel formed in this way, as long as the component kernels of the former efficiency dominate those of the latter. These results on the efficiency dominance of combining component kernels generalises the results on the efficiency dominance of combined chains introduced by Neal and Rosenthal from finite state spaces to general state spaces., Comment: 27 pages

Waveguide quantum electrodynamics, the study of atomic systems interacting with propagating electromagnetic fields, is a powerful platform for understanding the complex interplay between light and matter. Qubit control is an indispensable tool in this field, and most experiments have so far focused on narrowband electromagnetic waves that interact with qubits at specific frequencies. This interaction, however, changes significantly with fast, broadband pulses, as waveguide properties like dispersion affect the pulse evolution and its impact on the qubit. Here, we use dispersion to achieve spatial addressing of superconducting qubits separated by a sub-wavelength distance within a microwave waveguide. This novel approach relies on a self-focusing effect to create a position-dependent interaction between the pulse and the qubits. This experiment emphasizes the importance of dispersion in the design and analysis of quantum experiments, and offers new avenues for the rapid control of quantum states.

Experimental measurements of plasma and neutral profiles across the pedestal are used in conjunction with 2D edge modeling to examine pedestal stiffness in Alcator C-Mod H-mode plasmas. Experiments on Alcator C-Mod observed pedestal degradation and loss in confinement below a critical value of net power crossing the separatrix, $P_\mathrm{net} = P_\mathrm{net}^\mathrm{crit} \approx 2.3$ MW. New analysis of ionization and particle flux profiles reveal saturation of the pedestal electron density, $n_{e}^\mathrm{ped}$ despite continuous increases in ionization throughout the pedestal, inversely related to $P_\mathrm{net}$. A limit to the pedestal $\nabla n_{e}$ emerges as the particle flux, $\Gamma_{D}$ continues to grow, implying increases in the effective particle diffusivity, $D_\mathrm{eff}$. This is well-correlated with the separatrix collisionality, $\nu^{*}_\mathrm{sep}$ and a turbulence control parameter, $\alpha_{t}$, implying a possible transition in type of turbulence. The transition is well correlated with the experimentally observed value of $P_\mathrm{net}^\mathrm{crit}$. SOLPS-ITER modeling is performed for select discharges from the power scan, constrained with experimental electron and neutral densities, measured at the outer midplane. The modeling confirms general growth in $D_\mathrm{eff}$, consistent with experimental findings, and additionally suggests even larger growth in $\chi_{e}$ at the same $P_\mathrm{net}^\mathrm{crit}$.

This article develops general conditions for weak convergence of adaptive Markov chain Monte Carlo processes and is shown to imply a weak law of large numbers for bounded Lipschitz continuous functions. This allows an estimation theory for adaptive Markov chain Monte Carlo where previously developed theory in total variation may fail or be difficult to establish. Extensions of weak convergence to general Wasserstein distances are established along with a weak law of large numbers for possibly unbounded Lipschitz functions. Applications are applied to auto-regressive processes in various settings, unadjusted Langevin processes, and adaptive Metropolis-Hastings.

In this paper, we initiate the study of constant dimension subspace codes restricted to Schubert varieties, which we call Schubert subspace codes. These codes have a very natural geometric description, as objects that we call intersecting sets with respect to a fixed subspace. We provide a geometric construction of maximum size constant dimension subspace codes in some Schubert varieties with the largest possible value for the minimum subspace distance. Finally, we generalize the problem to different values of the minimum distance.

The Transiting Exoplanet Survey Satellite (TESS) has discovered hundreds of new worlds, with TESS planet candidates now outnumbering the total number of confirmed planets from $\textit{Kepler}$. Owing to differences in survey design, TESS continues to provide planets that are better suited for subsequent follow-up studies, including mass measurement through radial velocity (RV) observations, compared to Kepler targets. In this work, we present the TESS-Keck Survey's (TKS) Mass Catalog: a uniform analysis of all TKS RV survey data which has resulted in mass constraints for 126 planets and candidate signals. This includes 58 mass measurements that have reached $\geq5\sigma$ precision. We confirm or validate 32 new planets from the TESS mission either by significant mass measurement (15) or statistical validation (17), and we find no evidence of likely false positives among our entire sample. This work also serves as a data release for all previously unpublished TKS survey data, including 9,204 RV measurements and associated activity indicators over our three year survey. We took the opportunity to assess the performance of our survey, and found that we achieved many of our goals including measuring the mass of 38 small ($<4R_{\oplus}$) planets, nearly achieving the TESS mission's basic science requirement. In addition, we evaluated the performance of the Automated Planet Finder (APF) as survey support and observed meaningful constraints on system parameters due to its more uniform phase coverage. Finally, we compared our measured masses to those predicted by commonly used mass-radius relations and investigated evidence of systematic bias., Comment: 51 pages (22 of text), 24 figures

Large Language Models (LLM) have become a popular approach for implementing Retrieval Augmented Generation (RAG) systems, and a significant amount of effort has been spent on building good models and metrics. In spite of increased recognition of the need for rigorous evaluation of RAG systems, few tools exist that go beyond the creation of model output and automatic calculation. We present InspectorRAGet, an introspection platform for RAG evaluation. InspectorRAGet allows the user to analyze aggregate and instance-level performance of RAG systems, using both human and algorithmic metrics as well as annotator quality. InspectorRAGet is suitable for multiple use cases and is available publicly to the community. The demo video is available at https://youtu.be/MJhe8QIXcEc

The next generation of soft electronics will expand to the third dimension. This will require the integration of mechanically-compliant three-dimensional functional structures with stretchable materials. This study demonstrates omnidirectional direct ink writing (DIW) of Poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) aerogels with tunable electrical and mechanical performance, which can be integrated with soft substrates. Several PEDOT:PSS hydrogels were formulated for DIW and freeze-dried directly on stretchable substrates to form integrated aerogels displaying high shape fidelity and minimal shrinkage. The effect of additives and processing in the PEDOT:PSS hydro and aerogels morphology, and the link with their electromechanical properties was elucidated. This technology demonstrated 3D-structured stretchable interconnects and planar thermoelectric generators (TEGs) for skin electronics, as well as vertically-printed high aspect ratio thermoelectric pillars with a high ZT value of 3.2 10^-3 and ultra-low thermal conductivity of 0.065 W/(m K). Despite their comparatively low ZT, the aerogel pillars outpowered their dense counterparts in realistic energy harvesting scenarios where contact resistances cannot be ignored, and produced up to 26 nW/cm2 (corresponding to a gravimetric power density of 0.76 mW/kg) for a difference of temperature of 15 K. This work suggests promising advancements in soft and energy-efficiency electronic systems relevant to soft robotics and wearable.

Retrieval Augmented Generation (RAG) has become a popular application for large language models. It is preferable that successful RAG systems provide accurate answers that are supported by being grounded in a passage without any hallucinations. While considerable work is required for building a full RAG pipeline, being able to benchmark performance is also necessary. We present ClapNQ, a benchmark Long-form Question Answering dataset for the full RAG pipeline. ClapNQ includes long answers with grounded gold passages from Natural Questions (NQ) and a corpus to perform either retrieval, generation, or the full RAG pipeline. The ClapNQ answers are concise, 3x smaller than the full passage, and cohesive, with multiple pieces of the passage that are not contiguous. RAG models must adapt to these properties to be successful at ClapNQ. We present baseline experiments and analysis for ClapNQ that highlight areas where there is still significant room for improvement in grounded RAG. CLAPNQ is publicly available at https://github.com/primeqa/clapnq, Comment: 25 pages

The negatively charged tin-vacancy center in diamond (SnV$^-$) is an emerging platform for building the next generation of long-distance quantum networks. This is due to the SnV$^-$'s favorable optical and spin properties including bright emission, insensitivity to electronic noise, and long spin coherence times at temperatures above 1 Kelvin. Here, we demonstrate measurement of a single SnV$^-$ electronic spin with a single-shot readout fidelity of $87.4\%$, which can be further improved to $98.5\%$ by conditioning on multiple readouts. We show this performance is compatible with rapid microwave spin control, demonstrating that the trade-off between optical readout and spin control inherent to group-IV centers in diamond can be overcome for the SnV$^-$. Finally, we use weak quantum measurement to study measurement induced dephasing; this illuminates the fundamental interplay between measurement and decoherence in quantum mechanics, and makes use of the qubit's spin coherence as a metrological tool. Taken together, these results overcome an important hurdle in the development of the SnV$^-$ based quantum technologies, and in the process, develop techniques and understanding broadly applicable to the study of solid-state quantum emitters.

The Transiting Exoplanet Survey Satellite (TESS) continues to dramatically increase the number of known transiting exoplanets, and is optimal for monitoring bright stars amenable to radial velocity (RV) and atmospheric follow-up observations. TOI-1386 is a solar-type (G5V) star that was detected via TESS photometry to exhibit transit signatures in three sectors with a period of 25.84 days. We conducted follow-up RV observations using Keck/HIRES as part of the TESS-Keck Survey (TKS), collecting 64 RV measurements of TOI-1386 with the HIRES spectrograph over 2.5 years. Our combined fit of the TOI-1386 photometry and RV data confirm the planetary nature of the detected TESS signal, and provide a mass and radius for planet b of $0.148\pm0.019$ $M_J$ and $0.540\pm0.017$ $R_J$, respectively, marking TOI-1386 b as a warm sub-Saturn planet. Our RV data further reveal an additional outer companion, TOI-1386 c, with an estimated orbital period of 227.6 days and a minimum mass of $0.309\pm0.038$ $M_J$. The dynamical modeling of the system shows that the measured system architecture is long-term stable, although there may be substantial eccentricity oscillations of the inner planet due to the dynamical influence of the outer planet., Comment: Accepted for publication in The Astronomical Journal. 15 pages, 7 figures

The free distance of a convolutional code is a reliable indicator of its performance. However its computation is not an easy task. In this paper, we present some algorithms to compute the free distance with good efficiency that work for convolutional codes of all rates and over any field. Furthermore we discuss why an algorithm which is claimed to be very efficient is incorrect.

D-Nikud, a novel approach to Hebrew diacritization that integrates the strengths of LSTM networks and BERT-based (transformer) pre-trained model. Inspired by the methodologies employed in Nakdimon, we integrate it with the TavBERT pre-trained model, our system incorporates advanced architectural choices and diverse training data. Our experiments showcase state-of-the-art results on several benchmark datasets, with a particular emphasis on modern texts and more specified diacritization like gender.

The field of healthcare has increasingly turned its focus towards Large Language Models (LLMs) due to their remarkable performance. However, their performance in actual clinical applications has been underexplored. Traditional evaluations based on question-answering tasks don't fully capture the nuanced contexts. This gap highlights the need for more in-depth and practical assessments of LLMs in real-world healthcare settings. Objective: We sought to evaluate the performance of LLMs in the complex clinical context of adult critical care medicine using systematic and comprehensible analytic methods, including clinician annotation and adjudication. Methods: We investigated the performance of three general LLMs in understanding and processing real-world clinical notes. Concepts from 150 clinical notes were identified by MetaMap and then labeled by 9 clinicians. Each LLM's proficiency was evaluated by identifying the temporality and negation of these concepts using different prompts for an in-depth analysis. Results: GPT-4 showed overall superior performance compared to other LLMs. In contrast, both GPT-3.5 and text-davinci-003 exhibit enhanced performance when the appropriate prompting strategies are employed. The GPT family models have demonstrated considerable efficiency, evidenced by their cost-effectiveness and time-saving capabilities. Conclusion: A comprehensive qualitative performance evaluation framework for LLMs is developed and operationalized. This framework goes beyond singular performance aspects. With expert annotations, this methodology not only validates LLMs' capabilities in processing complex medical data but also establishes a benchmark for future LLM evaluations across specialized domains.

When employing Direct Volume Rendering (DVR) for visualizing volumetric scalar fields, classification is generally performed on a piecewise constant or piecewise linear approximation of the field on a viewing ray. Smoothed Particle Hydrodynamics (SPH) data sets define volumetric scalar fields as the sum of individual particle contributions, at highly varying spatial resolution. We present an approach for approximating SPH scalar fields along viewing rays with piece-wise polynomial functions of higher order. This is done by approximating each particle contribution individually and then efficiently summing the results, thus generating a higher-order representation of the field with a resolution adapting to the data resolution in the volume.

Most low-density parity-check (LDPC) code constructions are considered over finite fields. In this work, we focus on regular LDPC codes over integer residue rings and analyze their performance with respect to the Lee metric. Their error-correction performance is studied over two channel models, in the Lee metric. The first channel model is a discrete memoryless channel, whereas in the second channel model an error vector is drawn uniformly at random from all vectors of a fixed Lee weight. It is known that the two channel laws coincide in the asymptotic regime, meaning that their marginal distributions match. For both channel models, we derive upper bounds on the block error probability in terms of a random coding union bound as well as sphere packing bounds that make use of the marginal distribution of the considered channels. We estimate the decoding error probability of regular LDPC code ensembles over the channels using the marginal distribution and determining the expected Lee weight distribution of a random LDPC code over a finite integer ring. By means of density evolution and finite-length simulations, we estimate the error-correction performance of selected LDPC code ensembles under belief propagation decoding and a low-complexity symbol message passing decoding algorithm and compare the performances. The analysis developed in this paper may serve to design regular LDPC codes over integer residue rings for storage and cryptographic application.

Offensive language such as hate, abuse, and profanity (HAP) occurs in various content on the web. While previous work has mostly dealt with sentence level annotations, there have been a few recent attempts to identify offensive spans as well. We build upon this work and introduce Muted, a system to identify multilingual HAP content by displaying offensive arguments and their targets using heat maps to indicate their intensity. Muted can leverage any transformer-based HAP-classification model and its attention mechanism out-of-the-box to identify toxic spans, without further fine-tuning. In addition, we use the spaCy library to identify the specific targets and arguments for the words predicted by the attention heatmaps. We present the model's performance on identifying offensive spans and their targets in existing datasets and present new annotations on German text. Finally, we demonstrate our proposed visualization tool on multilingual inputs.

In January 2020, a workshop was held at EMBL-EBI (Hinxton, UK) to discuss data requirements for deposition and validation of cryoEM structures, with a focus on single-particle analysis. The meeting was attended by 45 experts in data processing, model building and refinement, validation, and archiving of such structures. This report describes the workshop's motivation and history, the topics discussed, and consensus recommendations resulting from the workshop. Some challenges for future methods-development efforts in this area are also highlighted, as is the implementation to date of some of the recommendations., Comment: Outcomes of a wwPDB/EMDB workshop on cryoEM data management, deposition and validation

Large language models (LLMs) have shown promise in safety-critical applications such as healthcare, yet the ability to quantify performance has lagged. An example of this challenge is in evaluating a summary of the patient's medical record. A resulting summary can enable the provider to get a high-level overview of the patient's health status quickly. Yet, a summary that omits important facts about the patient's record can produce a misleading picture. This can lead to negative consequences on medical decision-making. We propose MED-OMIT as a metric to explore this challenge. We focus on using provider-patient history conversations to generate a subjective (a summary of the patient's history) as a case study. We begin by discretizing facts from the dialogue and identifying which are omitted from the subjective. To determine which facts are clinically relevant, we measure the importance of each fact to a simulated differential diagnosis. We compare MED-OMIT's performance to that of clinical experts and find broad agreement We use MED-OMIT to evaluate LLM performance on subjective generation and find some LLMs (gpt-4 and llama-3.1-405b) work well with little effort, while others (e.g. Llama 2) perform worse., Comment: Accepted to ML4H 2024

We report on our effort to create a corpus dataset of different social context situations in an office setting for further disciplinary and interdisciplinary research in computer vision, psychology, and human-robot-interaction. For social robots to be able to behave appropriately, they need to be aware of the social context they act in. Consider, for example, a robot with the task to deliver a personal message to a person. If the person is arguing with an office mate at the time of message delivery, it might be more appropriate to delay playing the message as to respect the recipient's privacy and not to interfere with the current situation. This can only be done if the situation is classified correctly and in a second step if an appropriate behavior is chosen that fits the social situation. Our work aims to enable robots accomplishing the task of classifying social situations by creating a dataset composed of semantically annotated video scenes of office situations from television soap operas. The dataset can then serve as a basis for conducting research in both computer vision and human-robot interaction., Comment: In SCRITA 2023 Workshop Proceedings (arXiv:2311.05401) held in conjunction with 32nd IEEE International Conference on Robot & Human Interactive Communication, 28/08 - 31/08 2023, Busan (Korea)

Large language models (LLMs) have demonstrated their potential in social science research by emulating human perceptions and behaviors, a concept referred to as algorithmic fidelity. This study assesses the algorithmic fidelity and bias of LLMs by utilizing two nationally representative climate change surveys. The LLMs were conditioned on demographics and/or psychological covariates to simulate survey responses. The findings indicate that LLMs can effectively capture presidential voting behaviors but encounter challenges in accurately representing global warming perspectives when relevant covariates are not included. GPT-4 exhibits improved performance when conditioned on both demographics and covariates. However, disparities emerge in LLM estimations of the views of certain groups, with LLMs tending to underestimate worry about global warming among Black Americans. While highlighting the potential of LLMs to aid social science research, these results underscore the importance of meticulous conditioning, model selection, survey question format, and bias assessment when employing LLMs for survey simulation. Further investigation into prompt engineering and algorithm auditing is essential to harness the power of LLMs while addressing their inherent limitations., Comment: 34 pages, 6 figures, 1 table