Search

The ability of large language models (LLMs) to follow natural language instructions with human-level fluency suggests many opportunities in healthcare to reduce administrative burden and improve quality of care. However, evaluating LLMs on realistic text generation tasks for healthcare remains challenging. Existing question answering datasets for electronic health record (EHR) data fail to capture the complexity of information needs and documentation burdens experienced by clinicians. To address these challenges, we introduce MedAlign, a benchmark dataset of 983 natural language instructions for EHR data. MedAlign is curated by 15 clinicians (7 specialities), includes clinician-written reference responses for 303 instructions, and provides 276 longitudinal EHRs for grounding instruction-response pairs. We used MedAlign to evaluate 6 general domain LLMs, having clinicians rank the accuracy and quality of each LLM response. We found high error rates, ranging from 35% (GPT-4) to 68% (MPT-7B-Instruct), and an 8.3% drop in accuracy moving from 32k to 2k context lengths for GPT-4. Finally, we report correlations between clinician rankings and automated natural language generation metrics as a way to rank LLMs without human review. We make MedAlign available under a research data use agreement to enable LLM evaluations on tasks aligned with clinician needs and preferences.

SN 1987A was an unusual hydrogen-rich core-collapse supernova originating from a blue supergiant star. Similar blue supergiant explosions remain a small family of events, and are broadly characterized by their long rises to peak. The Zwicky Transient Facility (ZTF) Census of the Local Universe (CLU) experiment aims to construct a spectroscopically complete sample of transients occurring in galaxies from the CLU galaxy catalog. We identify 13 long-rising (>40 days) Type II supernovae from the volume-limited CLU experiment during a 3.5 year period from June 2018 to December 2021, approximately doubling the previously known number of these events. We present photometric and spectroscopic data of these 13 events, finding peak r-band absolute magnitudes ranging from -15.6 to -17.5 mag and the tentative detection of Ba II lines in 9 events. Using our CLU sample of events, we derive a long-rising Type II supernova rate of $1.37^{+0.26}_{-0.30}\times10^{-6}$ Mpc$^{-3}$ yr$^{-1}$, $\approx$1.4% of the total core-collapse supernova rate. This is the first volumetric rate of these events estimated from a large, systematic, volume-limited experiment., Comment: 33 pages, 18 figures, 5 tables. Updated to ApJ accepted version

As an essential prerequisite task in image-based plant phenotyping, leaf segmentation has garnered increasing attention in recent years. While self-supervised learning is emerging as an effective alternative to various computer vision tasks, its adaptation for image-based plant phenotyping remains rather unexplored. In this work, we present a self-supervised leaf segmentation framework consisting of a self-supervised semantic segmentation model, a color-based leaf segmentation algorithm, and a self-supervised color correction model. The self-supervised semantic segmentation model groups the semantically similar pixels by iteratively referring to the self-contained information, allowing the pixels of the same semantic object to be jointly considered by the color-based leaf segmentation algorithm for identifying the leaf regions. Additionally, we propose to use a self-supervised color correction model for images taken under complex illumination conditions. Experimental results on datasets of different plant species demonstrate the potential of the proposed self-supervised framework in achieving effective and generalizable leaf segmentation.

SPIRITS--the SPitzer InfraRed Intensive Transients Survey--searched for luminous infrared (IR) transients and variables in nearly 200 nearby galaxies from 2014 to 2019, using the warm Spitzer telescope at 3.6 and 4.5 microns. Among the SPIRITS variables are IR-bright objects that are undetected in ground-based optical surveys. We classify them as (1) transients, (2) periodic variables, and (3) irregular variables. The transients include "SPRITE"s (eSPecially Red Intermediate-luminosity Transient Events), having maximum luminosities fainter than supernovae, red IR colors, and a wide range of outburst durations (days to years). Here we report deep optical and near-IR imaging with the Hubble Space Telescope (HST) of 21 SPIRITS variables. They were initially considered SPRITE transients, but many eventually proved instead to be periodic or irregular variables as more data were collected. HST images show most of these cool and dusty variables are associated with star-forming regions in late-type galaxies, implying an origin in massive stars. Two SPRITEs lacked optical progenitors in deep pre-outburst HST images; however, one was detected during eruption at J and H, indicating a dusty object with an effective temperature of ~1050 K. One faint SPRITE turned out to be a dusty classical nova. About half the HST targets proved to be periodic variables, with pulsation periods of 670-2160 days; they are likely dusty asymptotic-giant-branch (AGB) stars with masses of ~5-10 Msun. A few of them were warm enough to be detected in deep HST frames, but most are too cool. Out of six irregular variables, two were red supergiants with optical counterparts in HST images; four were too enshrouded for HST detection., Comment: Accepted by Astrophysical Journal

We present a search for extragalactic fast blue optical transients (FBOTs) during Phase I of the Zwicky Transient Facility (ZTF). We identify 38 candidates with durations above half-maximum light 1 d < t1/2 < 12 d, of which 28 have blue (g-r<-0.2 mag) colors at peak light. Of the 38 transients (28 FBOTs), 19 (13) can be spectroscopically classified as core-collapse supernovae (SNe): 11 (8) H- or He-rich (Type II/IIb/Ib) SNe, 6 (4) interacting (Type IIn/Ibn) SNe, and 2 (1) H&He-poor (Type Ic/Ic-BL) SNe. Two FBOTs (published previously) had high-S/N predominantly featureless spectra and luminous radio emission: AT2018lug and AT2020xnd. Seven (five) did not have a definitive classification: AT 2020bdh showed tentative broad H$\alpha$ in emission, and AT 2020bot showed unidentified broad features and was 10 kpc offset from the center of an early-type galaxy. Ten (six) have no spectroscopic observations or redshift measurements. We present multiwavelength (radio, millimeter, and/or X-ray) observations for five FBOTs (three Type Ibn, one Type IIn/Ibn, one Type IIb). Additionally, we search radio-survey (VLA and ASKAP) data to set limits on the presence of radio emission for 22 of the transients. All X-ray and radio observations resulted in non-detections; we rule out AT2018cow-like X-ray and radio behavior for five FBOTs and more luminous emission (such as that seen in the Camel) for four additional FBOTs. We conclude that exotic transients similar to AT2018cow, the Koala, and the Camel represent a rare subset of FBOTs, and use ZTF's SN classification experiments to measure the rate to be at most 0.1% of the local core-collapse SN rate., Comment: Replaced following peer-review process. 46 pages, 20 figures. Accepted for publication in ApJ

The nova rate in the Milky Way remains largely uncertain, despite its vital importance in constraining models of Galactic chemical evolution as well as understanding progenitor channels for Type Ia supernovae. The rate has been previously estimated in the range of $\approx10-300$ yr$^{-1}$, either based on extrapolations from a handful of very bright optical novae or the nova rates in nearby galaxies; both methods are subject to debatable assumptions. The total discovery rate of optical novae remains much smaller ($\approx5-10$ yr$^{-1}$) than these estimates, even with the advent of all-sky optical time domain surveys. Here, we present a systematic sample of 12 spectroscopically confirmed Galactic novae detected in the first 17 months of Palomar Gattini-IR (PGIR), a wide-field near-infrared time domain survey. Operating in $J$-band ($\approx1.2$ $\mu$m) that is relatively immune to dust extinction, the extinction distribution of the PGIR sample is highly skewed to large extinction values ($> 50$% of events obscured by $A_V\gtrsim5$ mag). Using recent estimates for the distribution of mass and dust in the Galaxy, we show that the observed extinction distribution of the PGIR sample is commensurate with that expected from dust models. The PGIR extinction distribution is inconsistent with that reported in previous optical searches (null hypothesis probability $< 0.01$%), suggesting that a large population of highly obscured novae have been systematically missed in previous optical searches. We perform the first quantitative simulation of a $3\pi$ time domain survey to estimate the Galactic nova rate using PGIR, and derive a rate of $\approx 46.0^{+12.5}_{-12.4}$ yr$^{-1}$. Our results suggest that all-sky near-infrared time-domain surveys are well poised to uncover the Galactic nova population., Comment: 25 pages, 12 figures. Submitted to ApJ

We develop and evaluate a deep learning algorithm to classify multiple catheters on neonatal chest and abdominal radiographs. A convolutional neural network (CNN) was trained using a dataset of 777 neonatal chest and abdominal radiographs, with a split of 81%-9%-10% for training-validation-testing, respectively. We employed ResNet-50 (a CNN), pre-trained on ImageNet. Ground truth labelling was limited to tagging each image to indicate the presence or absence of endotracheal tubes (ETTs), nasogastric tubes (NGTs), and umbilical arterial and venous catheters (UACs, UVCs). The data set included 561 images containing 2 or more catheters, 167 images with only one, and 49 with none. Performance was measured with average precision (AP), calculated from the area under the precision-recall curve. On our test data, the algorithm achieved an overall AP (95% confidence interval) of 0.977 (0.679-0.999) for NGTs, 0.989 (0.751-1.000) for ETTs, 0.979 (0.873-0.997) for UACs, and 0.937 (0.785-0.984) for UVCs. Performance was similar for the set of 58 test images consisting of 2 or more catheters, with an AP of 0.975 (0.255-1.000) for NGTs, 0.997 (0.009-1.000) for ETTs, 0.981 (0.797-0.998) for UACs, and 0.937 (0.689-0.990) for UVCs. Our network thus achieves strong performance in the simultaneous detection of these four catheter types. Radiologists may use such an algorithm as a time-saving mechanism to automate reporting of catheters on radiographs., Comment: 10 pages, 5 figures (+1 suppl.), 2 tables (+2 suppl.). Submitted to Journal of Digital Imaging

Interaction-powered supernovae (SNe) explode within an optically-thick circumstellar medium (CSM) that could be ejected during eruptive events. To identify and characterize such pre-explosion outbursts we produce forced-photometry light curves for 196 interacting SNe, mostly of Type IIn, detected by the Zwicky Transient Facility between early 2018 and June 2020. Extensive tests demonstrate that we only expect a few false detections among the 70,000 analyzed pre-explosion images after applying quality cuts and bias corrections. We detect precursor eruptions prior to 18 Type IIn SNe and prior to the Type Ibn SN2019uo. Precursors become brighter and more frequent in the last months before the SN and month-long outbursts brighter than magnitude -13 occur prior to 25% (5 - 69%, 95% confidence range) of all Type IIn SNe within the final three months before the explosion. With radiative energies of up to $10^{49}\,\text{erg}$, precursors could eject $\sim1\,\text{M}_\odot$ of material. Nevertheless, SNe with detected precursors are not significantly more luminous than other SNe IIn and the characteristic narrow hydrogen lines in their spectra typically originate from earlier, undetected mass-loss events. The long precursor durations require ongoing energy injection and they could, for example, be powered by interaction or by a continuum-driven wind. Instabilities during the neon and oxygen burning phases are predicted to launch precursors in the final years to months before the explosion; however, the brightest precursor is 100 times more energetic than anticipated.

Several thousand core-collapse supernovae (CCSNe) of different flavors have been discovered so far. However, identifying their progenitors has remained an outstanding open question in astrophysics. Studies of SN host galaxies have proven to be powerful in providing constraints on the progenitor populations. In this paper, we present all CCSNe detected between 2009 and 2017 by the Palomar Transient Factory. This sample includes 888 SNe of 12 distinct classes out to redshift $z\approx1$. We present the photometric properties of their host galaxies from the far-ultraviolet to the mid-infrared and model the host-galaxy spectral energy distributions to derive physical properties. The galaxy mass functions of Type Ic, Ib, IIb, II, and IIn SNe ranges from $10^{5}$ to $10^{11.5}~M_\odot$, probing the entire mass range of star-forming galaxies down to the least-massive star-forming galaxies known. Moreover, the galaxy mass distributions are consistent with models of star-formation-weighted mass functions. Regular CCSNe are hence direct tracers of star formation. Small but notable differences exist between some of the SN classes. Type Ib/c SNe prefer galaxies with slightly higher masses (i.e., higher metallicities) and star-formation rates than Type IIb and II SNe. These differences are less pronounced than previously thought. H-poor SLSNe and SNe~Ic-BL are scarce in galaxies above $10^{10}~M_\odot$. Their progenitors require environments with metallicities of $<0.4$ and $<1$ solar, respectively. In addition, the hosts of H-poor SLSNe are dominated by a younger stellar population than all other classes of CCSNe. Our findings corroborate the notion that low-metallicity \textit{and} young age play an important role in the formation of SLSN progenitors., Comment: 24 pages main text, 14 figures, 9 Tables, catalogue available at http://www.github.com/steveschulze/PTF

(Abridged) Using the Zwicky Transient Facility alert stream, we are conducting a large campaign to spectroscopically classify all transients occurring in galaxies in the Census of the Local Universe (CLU) catalog. The aim of the experiment is to construct a spectroscopically complete, volume-limited sample of transients coincident within 100" of CLU galaxies out to 200 Mpc, and to a depth of 20 mag. We describe the survey design and spectroscopic completeness from the first 16 months of operations. We present results from a systematic search for Calcium rich gap transients in the sample of 22 low luminosity (peak absolute magnitude $M > -17$), hydrogen poor events found in the experiment (out of 754 spectroscopically classified SNe). We report the detection of eight Calcium rich gap transients, and constrain their volumetric rate to be at least $\approx 15\pm5$% of the SN Ia rate. Combining this sample with ten events from the literature, we find a likely continuum of spectroscopic properties ranging from events with SN Ia-like features (Ca-Ia objects) to SN Ib/c-like features (Ca-Ib/c objects) at peak light. Within the Ca-Ib/c events, we find two populations of events distinguished by their red ($g - r \approx 1.5$ mag) or green ($g - r \approx 0.5$ mag) spectral colors at $r$-band peak, wherein redder events show strong line blanketing signatures, slower light curves, weaker He lines and lower [Ca II]/[O I] in the nebular phase. Together, we find that the spectroscopic continuum, volumetric rates and striking old environments are consistent with the explosive burning of He shells on low mass white dwarfs. We posit that Ca-Ia and red Ca-Ib/c objects are consistent with the double detonation of He shells with high He burning efficiency, while green Ca-Ib/c objects could arise from less efficient He burning scenarios such as detonations in low density He shells or He shell deflagrations., Comment: 30 figures, 6 tables. Submitted to ApJ

M31-2015-LRN is a likely stellar merger discovered in the Andromeda Galaxy in 2015. We present new optical to mid-infrared photometry and optical spectroscopy for this event. Archival data shows that the source started to brighten $\sim$2 years before the nova event. During this precursor phase, the source brightened by $\sim$3 mag. The lightcurve at 6 and 1.5 months before the main outburst may show periodicity, with periods of 16$\pm$0.3 and 28.1$\pm$1.4 days respectively. This complex emission may be explained by runaway mass loss from the system after the binary undergoes Roche-lobe overflow, leading the system to coalesce in tens of orbital periods. While the progenitor spectral energy distribution shows no evidence of pre-existing warm dust in system, the remnant forms an optically thick dust shell at $\sim$4 months after the outburst peak. The optical depth of the shell increases dramatically after 1.5 years, suggesting the existence of shocks that enhance the dust formation process. We propose that the merger remnant is likely an inflated giant obscured by a cooling shell of gas with mass $\sim0.2$ M$_{\odot}$ ejected at the onset of the common envelope phase., Comment: 16 pages, 10 figures. Accepted for publication in MNRAS

Catheters are the second most common abnormal finding on radiographs. The position of catheters must be assessed on all radiographs, as serious complications can arise if catheters are malpositioned. However, due to the large number of radiographs performed each day, there can be substantial delays between the time a radiograph is performed and when it is interpreted by a radiologist. Computer-aided approaches hold the potential to assist in prioritizing radiographs with potentially malpositioned catheters for interpretation and automatically insert text indicating the placement of catheters in radiology reports, thereby improving radiologists' efficiency. After 50 years of research in computer-aided diagnosis, there is still a paucity of study in this area. With the development of deep learning approaches, the problem of catheter assessment is far more solvable. Therefore, we have performed a review of current algorithms and identified key challenges in building a reliable computer-aided diagnosis system for assessment of catheters on radiographs. This review may serve to further the development of machine learning approaches for this important use case., Comment: This manuscript has been accepted for publication in Radiology: Artificial Intelligence (https://pubs.rsna.org/journal/ai), which is published by the Radiological Society of North America (RSNA)

We present ZTF18aaqjovh (SN 2018bvw), a high-velocity ("broad-lined") stripped-envelope (Type Ic) supernova (Ic-BL SN) discovered in the Zwicky Transient Facility one-day cadence survey. ZTF18aaqjovh shares a number of features in common with engine-driven explosions: the photospheric velocity and the shape of the optical light curve are very similar to that of the Type Ic-BL SN 1998bw, which was associated with a low-luminosity gamma-ray burst (LLGRB) and had relativistic ejecta. However, the radio luminosity of ZTF18aaqjovh is almost two orders of magnitude fainter than that of SN 1998bw at the same velocity phase, and the shock velocity is at most mildly relativistic (v=0.06-0.4c). A search of high-energy catalogs reveals no compelling GRB counterpart to ZTF18aaqjovh, and the limit on the prompt GRB luminosity of $L_{\gamma,\mathrm{iso}} \approx 1.6 \times 10^{48}$ erg/sec excludes a classical GRB but not an LLGRB. Altogether, ZTF18aaqjovh represents another transition event between engine-driven SNe associated with GRBs and "ordinary" Ic-BL SNe., Comment: 18 pages, 8 figures. Accepted to the Astrophysical Journal on 2 March 2020. Revised from proofs on 4 Apr 2020

(Abridged) Palomar Gattini-IR is a new wide-field, near-infrared robotic time domain survey operating at Palomar Observatory. Using a 30 cm telescope mounted with a H2RG detector, Gattini-IR achieves a field of view of 25 sq. deg. with a pixel scale of 8.7" in J-band. Here, we describe the system design, survey operations, data processing system and on-sky performance of Palomar Gattini-IR. As a part of the nominal survey, Gattini-IR scans $\approx 7500$ square degrees of the sky every night to a median 5$\sigma$ depth of $15.7$ AB mag outside the Galactic plane. The survey covers $\approx 15000$ square degrees of the sky visible from Palomar with a median cadence of 2 days. A real-time data processing system produces stacked science images from dithered raw images taken on sky, together with PSF-fit source catalogs and transient candidates identified from subtractions within a median delay of $\approx 4$ hours from the time of observation. The calibrated data products achieve an astrometric accuracy (RMS) of $\approx 0.7$" with respect to Gaia DR2 for sources with S/N $> 10$, and better than $\approx 0.35$" for sources brighter than $\approx 12$ Vega mag. The photometric accuracy (RMS) achieved in the PSF-fit source catalogs is better than $\approx 3$% for sources brighter than $\approx 12$ Vega mag, as calibrated against the 2MASS catalog. With a field of view $\approx 40\times$ larger than any other existing near infrared imaging instrument, Gattini-IR is probing the reddest and dustiest transients in the local universe such as dust obscured supernovae in nearby galaxies, novae behind large columns of extinction within the galaxy, reddened micro-lensing events in the Galactic plane and variability from cool and dust obscured stars. We present results from transients and variables identified since the start of the commissioning period., Comment: 35 pages, 29 figures. Submitted to PASP

Transformers are critical assets in power systems and transformer failures can cause asset damage, customer outages, and safety concerns. Dominion Energy has a sophisticated monitoring process for the transformers. One of the most cost-efficient, convenient and practical transformer monitoring methods in industry is Dissolved Gas Analysis(DGA). Leveraging new technology, on-line transformer monitoring equipment is able to measure samples automatically. The challenges of unstable sampling measurements and contradicted analysis results for DGA are discussed in this paper. To provide further insight of transformer health and support a new transformer monitoring process in Dominion Energy, a DGA monitoring system is proposed. The DGA analysis methods used in the monitoring system are selected based on laboratory verification results from Dominion Energy. After derive the thresholds from IEEE standard, the solution of the proposed monitoring system and test results are presented. In the end, a historical transformer failure case in Dominion was analyzed and the results indicate the monitoring system can provide prescient information and sufficient supplemental report for making operational decisions., Comment: 9 pages

Recent studies on direct imaging of Type II core-collapse supernova progenitors indicate a possible threshold around $M_{\rm ZAMS}\sim 16-20$ M$_\odot$, where red supergiants with larger birth masses do not appear to result in supernova explosions and instead implode directly into a black hole. In this study we argue that it is not a coincidence that this threshold closely matches the critical transition of central Carbon burning in massive stars from the convective to radiative regime. In lighter stars, Carbon burns convectively in the center and result in compact final presupernova cores that are likely to result in explosions, while in heavier stars after the transition, it burns as a radiative flame and the stellar cores become significantly harder to explode. Using the KEPLER code we demonstrate the sensitivity of this transition to the rate of $^{12}$C$(\alpha,\gamma)^{16}$O reaction and the overshoot mixing efficiency, and we argue that the upper mass limit of exploding red supergiants could be employed to constrain uncertain input physics of massive stellar evolution calculations. The initial mass corresponding to the central Carbon burning transition range from 14 to 26 M$_\odot$ in recently published models from various groups and codes, and only a few are in agreement with the estimates inferred from direct imaging studies., Comment: submitted to MNRAS

We present a mid-infrared (IR) sample study of nearby ultraluminous X-ray sources (ULXs) using multi-epoch observations with the Infrared Array Camera (IRAC) on the Spitzer Space Telescope. Spitzer/IRAC observations taken after 2014 were obtained as part of the Spitzer Infrared Intensive Transients Survey (SPIRITS). Our sample includes 96 ULXs located within 10 Mpc. Of the 96~ULXs, 12 have candidate counterparts consistent with absolute mid-IR magnitudes of supergiants, and 16 counterparts exceeded the mid-IR brightness of single supergiants and are thus more consistent with star clusters or non-ULX background active galactic nuclei (AGN). The supergiant candidate counterparts exhibit a bi-modal color distribution in a Spitzer/IRAC color-magnitude diagram, where "red" and "blue" ULXs fall in IRAC colors $[3.6] - [4.5]\sim0.7$ and $[3.6] - [4.5]\sim0.0$, respectively. The mid-IR colors and absolute magnitudes of 4 "red" and 5 "blue" ULXs are consistent with that of supergiant B[e] (sgB[e]) and red supergiant (RSG) stars, respectively. While "blue", RSG-like mid-IR ULX counterparts likely host RSG mass donors, we propose the "red" counterparts are ULXs exhibiting the "B[e] phenomenon'' rather than hosts of sgB[e] mass donors. We show that the mid-IR excess from the "red" ULXs is likely due to thermal emission from circumstellar or circumbinary dust. Using dust as a probe for total mass, we estimate mass-loss rates of $\dot{M}\sim1\times10^{-4}$ M$_\odot$ yr$^{-1}$ in dust-forming outflows of red ULXs. Based on the transient mid-IR behavior and its relatively flat spectral index, $\alpha=-0.19\pm0.1$, we suggest that the mid-IR emission from Holmberg IX X-1 originates from a variable jet., Comment: 40 pages, 11 figures, accepted for publication in ApJ

We present the discovery of an optical transient (OT) in Messier 51, designated M51 OT2019-1 (also ZTF19aadyppr, AT 2019abn, ATLAS19bzl), by the Zwicky Transient Facility (ZTF). The OT rose over 15 days to an observed luminosity of $M_r=-13$ (${\nu}L_{\nu}=9\times10^6~L_{\odot}$), in the luminosity gap between novae and typical supernovae (SNe). Spectra during the outburst show a red continuum, Balmer emission with a velocity width of $\approx400$ km s$^{-1}$, Ca II and [Ca II] emission, and absorption features characteristic of an F-type supergiant. The spectra and multiband light curves are similar to the so-called "SN impostors" and intermediate-luminosity red transients (ILRTs). We directly identify the likely progenitor in archival Spitzer Space Telescope imaging with a $4.5~\mu$m luminosity of $M_{[4.5]}\approx-12.2$ and a $[3.6]-[4.5]$ color redder than 0.74 mag, similar to those of the prototype ILRTs SN 2008S and NGC 300 OT2008-1. Intensive monitoring of M51 with Spitzer further reveals evidence for variability of the progenitor candidate at [4.5] in the years before the OT. The progenitor is not detected in pre-outburst Hubble Space Telescope optical and near-IR images. The optical colors during outburst combined with spectroscopic temperature constraints imply a higher reddening of $E(B-V)\approx0.7$ mag and higher intrinsic luminosity of $M_r\approx-14.9$ (${\nu}L_{\nu}=5.3\times10^7~L_{\odot}$) near peak than seen in previous ILRT candidates. Moreover, the extinction estimate is higher on the rise than on the plateau, suggestive of an extended phase of circumstellar dust destruction. These results, enabled by the early discovery of M51 OT2019-1 and extensive pre-outburst archival coverage, offer new clues about the debated origins of ILRTs and may challenge the hypothesis that they arise from the electron-capture induced collapse of extreme asymptotic giant branch stars., Comment: 21 pages, 5 figures, published in ApJL