Search

This paper introduces LLM-jp, a cross-organizational project for the research and development of Japanese large language models (LLMs). LLM-jp aims to develop open-source and strong Japanese LLMs, and as of this writing, more than 1,500 participants from academia and industry are working together for this purpose. This paper presents the background of the establishment of LLM-jp, summaries of its activities, and technical reports on the LLMs developed by LLM-jp. For the latest activities, visit https://llm-jp.nii.ac.jp/en/.

Emotion plays a crucial role in human conversation. This paper underscores the significance of considering emotion in speech translation. We present the MELD-ST dataset for the emotion-aware speech translation task, comprising English-to-Japanese and English-to-German language pairs. Each language pair includes about 10,000 utterances annotated with emotion labels from the MELD dataset. Baseline experiments using the SeamlessM4T model on the dataset indicate that fine-tuning with emotion labels can enhance translation performance in some settings, highlighting the need for further research in emotion-aware speech translation systems., Comment: 9 pages. Accepted to ACL 2024 Findings. Dataset: https://huggingface.co/datasets/ku-nlp/MELD-ST

The creation of instruction data and evaluation benchmarks for serving Large language models often involves enormous human annotation. This issue becomes particularly pronounced when rapidly developing such resources for a non-English language like Japanese. Instead of following the popular practice of directly translating existing English resources into Japanese (e.g., Japanese-Alpaca), we propose an efficient self-instruct method based on GPT-4. We first translate a small amount of English instructions into Japanese and post-edit them to obtain native-level quality. GPT-4 then utilizes them as demonstrations to automatically generate Japanese instruction data. We also construct an evaluation benchmark containing 80 questions across 8 categories, using GPT-4 to automatically assess the response quality of LLMs without human references. The empirical results suggest that the models fine-tuned on our GPT-4 self-instruct data significantly outperformed the Japanese-Alpaca across all three base pre-trained models. Our GPT-4 self-instruct data allowed the LLaMA 13B model to defeat GPT-3.5 (Davinci-003) with a 54.37\% win-rate. The human evaluation exhibits the consistency between GPT-4's assessments and human preference. Our high-quality instruction data and evaluation benchmark have been released here., Comment: COLING 2024. Our code are available here: \href{https://github.com/hitoshizuku7/awesome-Ja-self-instruct}{self-instruct data} and \href{https://github.com/ku-nlp/ja-vicuna-qa-benchmark}{evaluation benchmark}

An ensemble of nitrogen-vacancy (NV) centers in diamond is a fascinating candidate to realize a sensitive magnetic field sensor. In particular, since the axes of the NV centers are distributed along four directions, a collection of measurement data from NV centers with different axes provides information on the vector components of a magnetic field. However, in the conventional approach, the low measurement contrast of NV centers limits the sensitivity of vector magnetic field sensing. Recently, to overcome this problem, multi-frequency control of the NV centers has been proposed. The key idea is that the four types of NV centers with different axes are simultaneously controlled by multi-frequency microwave pulses. Here, we demonstrate vector magnetic field sensing with an ensemble of NV centers in diamond via such multi-frequency control with pulsed-type measurements. We use Hahn echo pulses and extract information on the vector components of an applied AC magnetic field. We find that the sensitivity of diamond-based vector field sensing with multi-frequency control is better than that with single-frequency control for every vector component of a magnetic field., Comment: 5 pages, 5 figures

The problem-solving in automated theorem proving (ATP) can be interpreted as a search problem where the prover constructs a proof tree step by step. In this paper, we propose a deep reinforcement learning algorithm for proof search in intuitionistic propositional logic. The most significant challenge in the application of deep learning to the ATP is the absence of large, public theorem database. We, however, overcame this issue by applying a novel data augmentation procedure at each iteration of the reinforcement learning. We also improve the efficiency of the algorithm by representing the syntactic structure of formulas by a novel compact graph representation. Using the large volume of augmented data, we train highly accurate graph neural networks that approximate the value function for the set of the syntactic structures of formulas. Our method is also cost-efficient in terms of computational time. We will show that our prover outperforms Coq's $\texttt{tauto}$ tactic, a prover based on human-engineered heuristics. Within the specified time limit, our prover solved 84% of the theorems in a benchmark library, while $\texttt{tauto}$ was able to solve only 52%.

Psychiatric research has been hampered by an explanatory gap between psychiatric symptoms and their neural underpinnings, which has resulted in poor treatment outcomes. This situation has prompted us to shift from symptom-based diagnosis to data-driven diagnosis, aiming to redefine psychiatric disorders as disorders of neural circuitry. Promising candidates for data-driven diagnosis include resting-state functional connectivity MRI (rs-fcMRI)-based biomarkers. Although biomarkers have been developed with the aim of diagnosing patients and predicting the efficacy of therapy, the focus has shifted to the identification of biomarkers that represent therapeutic targets, which would allow for more personalized treatment approaches. This type of biomarker (i.e., theranostic biomarker) is expected to elucidate the disease mechanism of psychiatric conditions and to offer an individualized neural circuit-based therapeutic target based on the neural cause of a condition. To this end, researchers have developed rs-fcMRI-based biomarkers and investigated a causal relationship between potential biomarkers and disease-specific behavior using functional MRI (fMRI)-based neurofeedback on functional connectivity. In this review, we introduce recent approach for creating a theranostic biomarker, which consists mainly of two parts: (i) developing an rs-fcMRI-based biomarker that can predict diagnosis and/or symptoms with high accuracy, and (ii) the introduction of a proof-of-concept study investigating the relationship between normalizing the biomarker and symptom changes using fMRI-based neurofeedback. In parallel with the introduction of recent studies, we review rs-fcMRI-based biomarker and fMRI-based neurofeedback, focusing on the technological improvements and limitations associated with clinical use., Comment: 46 pages, 5 figures

By focusing on melancholic features with biological homogeneity, this study aimed to identify a small number of critical functional connections (FCs) that were specific only to the melancholic type of MDD. On the resting-state fMRI data, classifiers were developed to differentiate MDD patients from healthy controls (HCs). The classification accuracy was improved from 50 % (93 MDD and 93 HCs) to 70% (66 melancholic MDD and 66 HCs), when we specifically focused on the melancholic MDD with moderate or severer level of depressive symptoms. It showed 65% accuracy for the independent validation cohort. The biomarker score distribution showed improvements with escitalopram treatments, and also showed significant correlations with depression symptom scores. This classifier was specific to melancholic MDD, and it did not generalize in other mental disorders including autism spectrum disorder (ASD, 54% accuracy) and schizophrenia spectrum disorder (SSD, 45% accuracy). Among the identified 12 FCs from 9,316 FCs between whole brain anatomical node pairs, the left DLPFC / IFG region, which has most commonly been targeted for depression treatments, and its functional connections between Precuneus / PCC, and between right DLPFC / SMA areas had the highest contributions. Given the heterogeneity of the MDD, focusing on the melancholic features is the key to achieve high classification accuracy. The identified FCs specifically predicted the melancholic MDD and associated with subjective depressive symptoms. These results suggested key FCs of melancholic depression, and open doors to novel treatments targeting these regions in the future.

Obsessive-compulsive disorder (OCD) is a common psychiatric disorder with a lifetime prevalence of 2-3 percent. Recently, brain activity in the resting state is gathering attention as a new means of exploring altered functional connectivity in psychiatric disorders. Although previous resting-state functional magnetic resonance imaging studies investigated neurobiological abnormalities of patients with OCD, there are concerns that should be addressed. One concern is the validity of the hypothesis employed. Most studies used seed-based analysis of the fronto-striatal circuit, despite the potential for abnormalities in other regions. A hypothesis-free study is a promising approach in such a case, while it requires researchers to handle a dataset with large dimensions. Another concern is the reliability of biomarkers derived from a single dataset, which may be influenced by cohort-specific features. Here, by employing a recently developed machine-learning algorithm to avoid these concerns, we identified the first OCD biomarker that is generalized to an external dataset. We also demonstrated that the functional connectivities that contributed to the classification were widely distributed rather than locally constrained. Our generalizable classifier has the potential not only to deepen our understanding of the abnormal neural substrates of OCD but also to find use in clinical applications., Comment: 47 pages, 3 figures

The most widely used Galactic extinction map (Schlegel, Finkbeiner, & Davis 1998, SFD) is constructed assuming that the observed FIR fluxes entirely come from the Galactic dust. According to the earlier suggestion by Yahata et al. (2007), we consider how far-infrared (FIR) emission of galaxies affects the SFD map. We first compute the surface number density of SDSS DR7 galaxies as a function of the $r$-band extinction, $A_{r,\rm SFD}$. We confirm that the surface densities of those galaxies positively correlate with $A_{r,\rm SFD}$ for $A_{r,\rm SFD}<0.1$, as first discovered by Yahata et al. (2007) for SDSS DR4 galaxies. Next we construct an analytic model to compute the surface density of galaxies taking account of the contamination of their FIR emission. We adopt a log-normal probability distribution for the ratio of $100\mu {\rm m}$ and $r$-band luminosities of each galaxy, $y \equiv (\nu L)_{100\mu {\rm m}}/(\nu L)_r$. Then we search for the mean and r.m.s values of $y$ that fit the observed anomaly using the analytic model. The required values to reproduce the anomaly are roughly consistent with those measured from the stacking analysis of SDSS galaxies (Kashiwagi, Yahata, & Suto 2013). Due to the limitation of our statistical modeling, we are not yet able to remove the FIR contamination of galaxies from the extinction map. Nevertheless the agreement with the model prediction suggests that the FIR emission of galaxies is mainly responsible for the observed anomaly. While the corresponding systematic error in the Galactic extinction map is 0.1 to 1mmag, it is directly correlated with galaxy clustering, and thus needs to be carefully examined in precision cosmology., Comment: 20 pages, 12 figures, accepted for publication in ApJ

We have performed stacking image analyses of galaxies over the Galactic extinction map constructed by Schlegel, Finkbeiner & Davis (1998). We select ~10^7 galaxies in total from the Sloan Digital Sky Survey (SDSS) DR7 photometric catalog. We detect clear signatures of the enhancement of the extinction in r-band, $\Delta A_r$, around galaxies, indicating that the extinction map is contaminated by their FIR (far infrared) emission. The average amplitude of the contamination per galaxy is well fitted to $\Delta A_r(m_r) = 0.64 \times 10^{0.17(18-m_r)}$ [mmag]. While this value is very small, it is directly associated with galaxies and may have a systematic effect on galaxy statistics. Indeed this correlated contamination leads to a relatively large anomaly of galaxy surface number densities against the SFD extinction A_SFD discovered by Yahata et al. (2007). We model the radial profiles of stacked galaxy images, and find that the FIR signal around each galaxy does not originate from the central galaxy alone, but is dominated by the contributions of nearby galaxies via galaxy angular clustering. The separation of the single galaxy and the clustering terms enables us to infer the statistical relation of the FIR and r-band fluxes of galaxies and also to probe the flux-weighted cross-correlation of galaxies, down to the magnitudes that are difficult to probe directly for individual objects. We repeat the same stacking analysis for SDSS DR6 photometric quasars and discovered the similar signatures but with weaker amplitudes. The implications of the present results for galaxy and quasar statistics and for correction to the Galactic extinction map are briefly discussed., Comment: 11 pages, 19 figures, PASJ, 2013, vol65, No.3, in press

The peculiar motion of the Earth causes a dipole anisotropy modulation in the distant galaxy distribution due to the aberration effect. However, the amplitude and angular direction of the effect is not necessarily the same as those of the cosmic microwave background (CMB) dipole anisotropy due to the growth of cosmic structures. In other words exploring the aberration effect may give us a clue to the horizon-scale physics perhaps related to the cosmic acceleration. In this paper we develop a method to explore the dipole angular modulation from the pixelized galaxy data on the sky properly taking into account the covariances due to the shot noise and the intrinsic galaxy clustering contamination as well as the partial sky coverage. We applied the method to the galaxy catalogs constructed from the Sloan Digital Sky Survey (SDSS) Data Release 6 data. After constructing the four galaxy catalogs that are different in the ranges of magnitudes and photometric redshifts to study possible systematics, we found that the most robust sample against systematics indicates no dipole anisotropy in the galaxy distribution. This finding is consistent with the expectation from the concordance Lambda-dominated cold dark matter model. Finally we argue that an almost full-sky galaxy survey such as LSST may allow for a significant detection of the aberration effect of the CMB dipole having the precision of constraining the angular direction to ~ 20 degrees in radius. Assuming a hypothetical LSST galaxy survey, we find that this method can confirm or reject the result implied from a stacked analysis of the kinetic Sunyaev-Zel'dovich effect of X-ray luminous clusters in Kashlinsky et al. (2008,2009) if the implied cosmic bulk flow is not extended out to the horizon., Comment: 20 pages, 11 figures; 24 pages, added a couple of references and 2 figures. Revised version in response to the referee's comments. Resubmitted to Phys. Rev. D

We investigate correlation between the arrival directions of ultra-high-energy cosmic rays (UHECRs) and the large-scale structure (LSS) of the Universe by using statistical quantities which can find the angular scale of the correlation. The Infrared Astronomical Satellite Point Source Redshift Survey (IRAS PSCz) catalog of galaxies is adopted for LSS. We find a positive correlation of the highest energy events detected by the Pierre Auger Observatory (PAO) with the IRAS galaxies inside $z=0.018$ within the angular scale of $\sim 15^{\circ}$. This positive correlation observed in the southern sky implies that a significant fraction of the highest energy events comes from nearby extragalactic objects. We also analyze the data of the Akeno Giant Air Shower Array (AGASA) which observed the northern hemisphere, but the obvious signals of positive correlation with the galaxy distribution are not found. Since the exposure of the AGASA is smaller than the PAO, the cross-correlation in the northern sky should be tested using a larger number of events detected in the future. We also discuss the correlation using the all-sky combined data sets of both the PAO and AGASA, and find a significant correlation within $\sim 8^{\circ}$. These angular scales can constrain several models of intergalactic magnetic field. These cross-correlation signals can be well reproduced by a source model in which the distribution of UHECR sources is related to the IRAS galaxies., Comment: 21 pages,7 figures

We use a series of cosmological N-body simulations and various analytic models to study the evolution of the matter power spectrum in real space in a \Lambda Cold Dark Matter universe. We compare the results of N-body simulations against three analytical model predictions; standard perturbation theory, renormalized perturbation theory, and the closure approximation. We include the effects from finite simulation box size in the comparison. We determine the values of the maximum wavenumbers, k^{lim}_{1%} and k^{lim}_{3%}, below which the analytic models and the simulation results agree to within 1 and 3 percent, respectively. We then provide a simple empirical function which describes the convergence regime determined by comparison between our simulations and the analytical models. We find that if we use the Fourier modes within the convergence regime alone, the characteristic scale of baryon acoustic oscillations can be determined within 1% accuracy from future surveys with a volume of a few h^{-3}Gpc^3 at z\sim1 or z\sim3 in the absence of any systematic distortion of the power spectrum., Comment: 23 pages, 14 figures, accepted by PASJ

We critically examine how well the evolution of large-scale density perturbations is followed in cosmological $N$-body simulations. We first run a large volume simulation and perform a mode-by-mode analysis in three-dimensional Fourier space. We show that the growth of large-scale fluctuations significantly deviates from linear theory predictions. The deviations are caused by {\it nonlinear} coupling with a small number of modes at largest scales owing to finiteness of the simulation volume. We then develop an analytic model based on second-order perturbation theory to quantify the effect. Our model accurately reproduces the simulation results. For a single realization, the second-order effect appears typically as ``zig-zag'' patterns around the linear-theory prediction, which imprints artificial ``oscillations'' that lie on the real baryon-acoustic oscillations. Although an ensemble average of a number of realizations approaches the linear theory prediction, the dispersions of the realizations remain large even for a large simulation volume of several hundred megaparsecs on a side. For the standard $\Lambda$CDM model, the deviations from linear growth rate are as large as 10 percent for a simulation volume with $L = 500h^{-1}$Mpc and for a bin width in wavenumber of $\Delta k = 0.005h$Mpc$^{-1}$, which are comparable to the intrinsic variance of Gaussian random realizations. We find that the dispersions scales as $\propto L^{-3/2} \Delta k^{-1/2}$ and that the mean dispersion amplitude can be made smaller than a percent only if we use a very large volume of $L > 2h^{-1}$Gpc. The finite box size effect needs to be appropriately taken into account when interpreting results from large-scale structure simulations for future dark energy surveys using baryon acoustic oscillations., Comment: references added, minor changes

An acoustic oscillation of the primeval photon-baryon fluid around the decoupling time imprints a characteristic scale in the galaxy distribution today, known as the baryon acoustic oscillation (BAO) scale. Several on-going and/or future galaxy surveys aim at detecting and precisely determining the BAO scale so as to trace the expansion history of the universe. We consider nonlinear and redshift-space distortion effects on the shifts of the BAO scale in $k$-space using perturbation theory. The resulting shifts are indeed sensitive to different choices of the definition of the BAO scale, which needs to be kept in mind in the data analysis. We present a toy model to explain the physical behavior of the shifts. We find that the BAO scale defined as in Percival et al. (2007) indeed shows very small shifts ($\lesssim$ 1%) relative to the prediction in {\it linear theory} in real space. The shifts can be predicted accurately for scales where the perturbation theory is reliable., Comment: 21 pages, 9 figures, references and supplementary sections added, accepted for publication in PASJ

We consider nonlinear biasing models of galaxies with particular attention to a correlation between linear and quadratic biasing coefficients, b_1 and b_2. We first derive perturbative expressions for b_1 and b_2 in halo and peak biasing models. Then we compute power spectra and bispectra of dark matter particles and halos using N-body simulation data and of volume-limited subsamples of Sloan Digital Sky Survey (SDSS) galaxies, and determine their b_1 and b_2. We find that the values of those coefficients at linear regimes (k<0.2h/Mpc) are fairly insensitive to the redshift-space distortion and the survey volume shape. The resulting normalized amplitudes of bispectra, Q, for equilateral triangles, are insensitive to the values of b_1 implying that b_2 indeed correlates with b_1. The present results explain the previous finding of Kayo et al. (2004) for the hierarchical relation of three-point correlation functions of SDSS galaxies. While the relations between b_1 and b_2 are quantitatively different for specific biasing models, their approximately similar correlations indicate a fairly generic outcome of the biasing due to the gravity in primordial Gaussian density fields., Comment: 14 pages, 8 figures, accepted for publication in PASJ

We compare the most successful and widely used map of Galactic dust extinction, provided by Schlegel, Finkbeiner & Davis (1998; hereafter SFD), to the galaxy number counts in the Sloan Digital Sky Survey (SDSS) photometric/spectroscopic DR4 sample. We divide the SDSS survey area into 69 disjoint subregions according to the dust extinction provided by SFD and compare the surface number density of galaxies in each subregion. As expected, the galaxy surface number density decreases with increasing extinction but only for SFD extinction values above about 0.1 to 0.2 magnitudes (depending on the band). At lower values of the SFD extinction, we find that the sky surface density of galaxies increases with increasing extinction, precisely the opposite of the effect expected from Galactic dust. We suggest that the far infrared (FIR) brightness of the sky in regions of true low dust extinction is significantly ``contaminated'' by the FIR emission from background galaxies. We show that such an explanation is both qualitatively and quantitatively consistent with the available data. Based on this interpretation we conclude that systematic errors in the SFD extinction map due to extragalactic FIR emission are quite small, of order hundredths of a magnitude, but nevertheless statistically detectable. (Abridged), Comment: 23 pages, 14 figures, submitted to PASJ

We discuss how the baryon acoustic oscillation (BAO) signatures in the galaxy power spectrum can distinguish between modified gravity and the cosmological constant as the source of cosmic acceleration. To this end we consider a model characterized by a parameter $n$, which corresponds to the Dvali-Gabadadze-Porrati (DGP) model if $n=2$ and reduces to the standard spatially flat cosmological constant concordance model for $n$ equal to infinity. We find that the different expansion histories of the modified gravity models systematically shifts the peak positions of BAO. A preliminary analysis using the current SDSS LRG sample indicates that the original DGP model is disfavored unless the matter density parameter exceeds 0.3. The constraints will be strongly tightened with future spectroscopic samples of galaxies at high redshifts. We demonstrate that WFMOS, in collaboration with other surveys such as Planck, will powerfully constrain modified gravity alternatives to dark energy as the explanation of cosmic acceleration., Comment: Physical Review D, in press

We report the first result of the clustering analysis of Sloan Digital Sky Survey (SDSS) quasars. We compute the two-point correlation function (2PCF) of SDSS quasars in redshift space at $8h^{-1}{\rm Mpc} < s < 500h^{-1}{\rm Mpc}$, with particular attention to its baryonic signature. Our sample consists of 19986 quasars extracted from the SDSS Data Release 4 (DR4). The redshift range of the sample is $0.72 \le z \le 2.24$ (the mean redshift is $\bar z = 1.46$) and the reddening-corrected $i$-band apparent magnitude range is $15.0 \le m_{i,{\rm rc}} \le 19.1$. Due to the relatively low number density of the quasar sample, the bump in the power spectrum due to the baryon density, $\Omega_{\rm b}$, is not clearly visible. The effect of the baryon density is, however, to distort the overall shape of the 2PCF.The degree of distortion makes it an interesting alternate measure of the baryonic signature. Assuming a scale-independent linear bias and the spatially flat universe, i.e., $\Omega_{\rm b} + \Omega_{\rm d} + \Omega_\Lambda =1$, where $\Omega_{\rm d}$ and $\Omega_\Lambda$ denote the density parameters of dark matter and the cosmological constant, we combine the observed quasar 2PCF and the predicted matter 2PCF to put constraints on $\Omega_{\rm b}$ and $\Omega_\Lambda$. Our result is fitted as $0.80- 2.8\Omega_{\rm b} < \Omega_\Lambda < 0.90 - 1.4\Omega_{\rm b}$ at the 2$\sigma$ confidence level, which is consistent with results from other cosmological observations such as WMAP. (abridged), Comment: 26 pages, 12 figures, Accepted for publication in the PASJ

This paper describes the acquisition and analysis of imaging data for the identification of galaxies associated with z~4 damped Lya systems. We present deep BRI images of three fields known to contain four z~4 damped systems. We discuss the reduction and calibration of the data, detail the color criteria used to identify z~4 galaxies, and present a photometric redshift analysis to complement the color selection. We have found no galaxy candidates closer to the QSO than 7'' which could be responsible for the damped Lya systems. Assuming that at least one of the galaxies is not directly beneath the QSO, we set an upper limit on this damped Lya system of L < L*/4. Finally, we have established a web site to release these imaging data to the public., Comment: 12 pages, 6 embedded figures (3 color), 9 jpg figures. Higher quality ps versions of the images and the fits data are available at http://kingpin.ucsd.edu/~dlaimg, Accepted to the Astronomical Journal Jan 22, 2002

We present a calculation of the systematic component of the error budget in the photometric redshift technique. We make use of it to describe a simple technique that allows for the assignation of confidence limits to redshift measurements obtained through photometric methods. We show that our technique, through the calculation of a redshift probability function, gives complete information on the probable redshift of an object and its associated confidence intervals. This information can and must be used in the calculation of any observable quantity which makes use of the redshift., Comment: 6 pages, 5 figures, MNRAS style. Accepted for publication in MNRAS

We address the effects of cosmological surface brightness dimming on observations of faint galaxies by examining the distribution of "unobscured" star formation rate intensities versus redshift. We use the star formation rate intensity distribution function to assess the ultraviolet luminosity density versus redshift, based on our photometry and photometric redshift measurements of faint galaxies in the HDF and the HDF--S WFPC2 and NICMOS fields. We find that (1) previous measurements have missed a dominant fraction of the ultraviolet luminosity density of the universe at high redshifts by neglecting cosmological surface brightness dimming effects, which are important at redshifts larger than z = 2, (2) the incidence of the highest intensity star forming regions increases monotonically with redshift, and (3) the ultraviolet luminosity density plausibly increases monotonically with redshift through the highest redshifts observed. By measuring the spectrum of the luminosity density versus redshift, we also find that (4) previous measurements of the ultraviolet luminosity density at redshifts z < 2 must be reduced by a factor 2 to allow for the spectrum of the luminosity density between rest-frame wavelengths 1500 and 2800 A. And by comparing with observations of high-redshift damped Lyman-alpha absorption systems detected toward background QSOs, we further find that (5) the distribution of star formation rate intensities matches the distribution of neutral hydrogen column densities at redshifts z = 2 through 5, which establishes a quantitative connection between high-redshift galaxies and high column density gas and suggests that high-redshift damped Lyman-alpha absorption systems trace lower star formation rate intensity regions of the same galaxies detected in star light in the HDF and HDF--S., Comment: 28 pages, 9 figures; accepted for publication in the Astrophysical Journal

We report the first observations of the clustering environment of damped Lyman alpha absorption systems at z=4. Color selection and photometric redshifts were used to select 44 candidate Lyman-break galaxies brighter than I_AB=25.5 from deep BRI images of the 35 sq. arcmin field containing the quasar BR 0951-04. Multislit spectroscopy of 35 candidate galaxies was performed and 8 of these candidates have been confirmed as z>3.5 Lyman-break galaxies. With only BRI photometry, the photometric redshifts are quite accurate for the spectroscopically confirmed galaxies but have a high rate of misclassification due to color degeneracies between Lyman-break galaxies and low-redshift ellipticals. Both of the z>3.5 galaxies found within 15'' of the quasar line-of-sight appear to be causing absorption systems in the quasar spectrum. We use a battery of statistical tests to look for clustering in the redshift histogram of the z>3.5 galaxies but do not find measurable clustering of these Lyman-break galaxies with the damped Lyman alpha absorbers. With a larger sample of galaxies, our method should determine the cross-correlation between these objects, which probes the bias and hence the mass of the damped Lyman alpha absorbers., Comment: submitted to ApJ, revised version in response to referee's comments, one conclusion changed, 9 pages including 3 color figures

Observations of distant galaxies are important both for understanding how galaxies form and for probing the physical conditions of the universe at the earliest epochs. It is, however, extremely difficult to identify galaxies at redshift z>5, because these galaxies are faint and exhibit few spectral features. In a previous work, we presented observations that supported the identification of a galaxy at redshift z = 6.68 in a deep STIS field. Here we present new ground-based photometry of the galaxy. We find that the galaxy exhibits moderate detections of flux in the optical B and V images that are inconsistent with the expected absence of flux at wavelength shortward of the redshifted Lyman-alpha emission line of a galaxy at redshift z>5. In addition, the new broad-band imaging data not only show flux measurements of this galaxy that are incompatible with the previous STIS measurement, but also suggest a peculiar spectral energy distribution that cannot be fit with any galaxy spectral template at any redshift. We therefore conclude that the redshift identification of this galaxy remains undetermined., Comment: 9 pages, 2 figures; To appear in Nature (30 November 2000)

We present a comparison between the spectroscopic catalog of the HDF recently published by Cohen et al and the redshifts measured by our group for those objects using photometric techniques. We perform this comparison in order to characterize the errors associated to the photometric redshift technique. The spectroscopic sample includes over 140 objects in the HDF proper, representing the deepest, cleanest, most complete spectroscopic catalog ever compiled. We study each object for which our redshift and the one measured by Cohen et al disagree. In most cases the photometric evidence is strong enough to call for a careful review of the spectroscopic values, as they seem to be in error. It is possible to characterize the systematic errors associated to our technique, which when combined with the photometric errors allow us to obtain complete information on the redshift of each galaxy and its associated confidence interval, regardless of apparent magnitude. One of the main conclusions of this study is that, to date, all the redshifts from our published catalog that have been checked have been shown to be correct (within the stated confidence limits). This implies that our galaxy template set is a fair representation of the galaxy population at all redshifts (0

We present a catalog of photometry and photometric redshifts of 335 faint objects in the HDF-S NICMOS field. The analysis is based on (1) infrared images obtained with the Hubble Space Telescope (HST) using the Near Infrared Camera and Multi-Object Spectrograph (NICMOS) with the F110W, F160W, and F222M filters, (2) an optical image obtained with HST using the Space Telescope Imaging Spectrograph (STIS) with no filter, and (3) optical images obtained with the European Southern Observatory (ESO) Very Large Telescope (VLT) with U, B, V, R, and I filters. The primary utility of the catalog of photometric redshifts is as a survey of faint galaxies detected in the NICMOS F160W and F222M images. The sensitivity of the survey varies significantly with position, reaching a limiting depth of AB(16,000) ~ 28.7 and covering 1.01 arcmin^2 to AB(16,000) = 27 and 1.05 arcmin^2 to AB(16,000) = 26.5. The catalog of photometric redshifts identifies 21 galaxies (or 6% of the total) of redshift z > 5, 8 galaxies (or 2% of the total) of redshift z > 10, and 11 galaxies (or 3% of the total) of best-fit spectral type E/S0, of which 5 galaxies (or 1% of the total) are of redshift z > 1., Comment: 33 pages, 10 figures, accepted for publication in the Astrophysical Journal, August 1, 2000 issue

We report on some aspects of our efforts to establish properties of the extremely faint galaxy population by applying our photometric redshift technique to the HDF and HDF-S WFPC2 and NICMOS fields. We find that cosmological surface brightness dimming effects play a dominant role in setting what is observed at redshifts z > 2, that the comoving number density of high intrinsic surface brightness regions increases monotonically with increasing redshift, and that previous estimates neglect a significant or dominant fraction of the ultraviolet luminosity density of the universe due to surface brightness effects. The ultraviolet luminosity density of the universe plausibly increases monotonically with increasing redshift to redshifts beyond z = 5., Comment: 10 pages, 6 figures, includes style file, to appear in "The Hy-Redshift Universe: Galaxy Formation and Evolution at High Redshift" Conference Proceedings (editors Andrew J. Bunker & Wil J. M. van Breugel)

We report on some aspects of the current status of our efforts to establish properties of faint galaxies by applying our photometric redshift technique to faint galaxies in the HDF and HDF-S WFPC2 and NICMOS fields., Comment: 6 pages, 4 figures, to appear in "Photometric Redshifts and High Redshift Galaxies", eds. R. Weymann, L. Storrie-Lombardi, M. Sawicki & R. Brunner

We present an electronic catalog of infrared and optical photometry and photometric redshifts of 323 galaxies in the Hubble Deep Field South NICMOS field at http://www.ess.sunysb.edu/astro/hdfs/home.html. The analysis is based on infrared images obtained with the Hubble Space Telescope using the Near Infrared Camera and Multi-Object Spectrograph and the Space Telescope Imaging Spectrograph together with optical images obtained with the Very Large Telescope. The infrared and optical photometry is measured by means of a new quasi-optimal photometric technique that fits model spatial profiles of the galaxies determined by Pixon image reconstruction techniques to the images. In comparison with conventional methods, the new technique provides higher signal-to-noise-ratio measurements and accounts for uncertainty correlations between nearby, overlapping neighbors. The photometric redshifts are measured by means of our redshift likelihood technique, incorporating six spectrophotometric templates which, by comparison with spectroscopic redshifts of galaxies identified in the Hubble Deep Field North, are known to provide redshift measurements accurate to within an RMS relative uncertainty of (Delta z)/(1 + z) < 0.1 at all redshifts z < 6. The analysis reaches a peak H-band sensitivity threshold of AB(16000) = 28.3 and covers 1.02 acrmin^2 to AB(16000) = 27, 1.27 arcmin^2 to AB(16000) = 26, and 1.44 arcmin^2 to AB(16000) = 25. The analysis identifies galaxies at redshifts ranging from z near 0 through z greater than 10, including 17 galaxies of redshift 5 < z < 10 and five candidate galaxies of redshift z > 10., Comment: 1 page

We present new statistical study on the origin of Lyman-alpha absorption systems that are detected in QSO fields. The primary objective is to investigate environmental dependence of the relationship between galaxies and Lyman-alpha absorption systems. We find that galaxies exhibit no environmental dependence in giving rise to Lyman-alpha absorbers in their halos. Conversely, Lyman-alpha absorbers do not exhibit preference as to the environment where they reside., Comment: 4 pages, 1 figure, to appear in proceedings of the Xth Rencontres de Blois on "The Birth of Galaxies," LaTex style file included

Here we describe our attempts to establish statistically complete samples of very high redshift galaxies by obtaining photometric redshifts of galaxies in Medium Deep Survey (MDS) fields and photometric and spectroscopic redshifts of galaxies in very deep STIS slitless spectroscopy fields. On the basis of this analysis, we have identified galaxies of redshift z = 4.92 in an MDS field and of redshift z = 6.68 in a very deep STIS field., Comment: 4 pages, 2 figures, to appear in proceedings of the Xth Recontres de Blois, "The Birth of Galaxies," LaTeX style file included

We report new HST and ground-based observations of a damped Lyman-alpha absorption system toward the QSO 0850+4400. The redshift of the absorption system is z = 0.163770 and the neutral hydrogen column density of the absorption system is log N = 19.81 cm**-2. The absorption system is by far the lowest redshift confirmed damped Lyman-alpha absorption system yet identified, which provides an unprecedented opportunity to examine the nature, impact geometry, and kinematics of the absorbing galaxy in great detail. The observations indicate that the absorption system is remarkable in three respects: First, the absorption system is characterized by weak metal absorption lines and a low metal abundance, possibly less than 4% of the solar metal abundance. This cannot be explained as a consequence of dust, because the neutral hydrogen column density of the absorption system is far too low for obscuration by dust to introduce any significant selection effects. Second, the absorption system is associated with a moderate-luminosity early-type S0 galaxy, although the absorption may actually arise in one of several very faint galaxies detected very close to the QSO line of sight. Third, the absorbing material moves counter to the rotating galaxy disk, which rules out the possibility that the absorption arises in a thin or thick co-rotating gaseous disk. These results run contrary to the expectation that low-redshift damped Lyman-alpha absorption systems generally arise in the gas- and metal-rich inner parts of late-type spiral galaxies. We suggest instead that mounting evidence indicates that low-redshift galaxies of a variety of morphological types may contain significant quantities of low metal abundance gas at large galactocentric distances., Comment: 15 pages, LaTex, 4 figures, to be published in The Astronomical Journal

NASTRAN and substructuring techniques for structural analysis of complex space shuttle structures

Pre-processor, post-processor, and multiple level substructuring routines for data management in support of NASTRAN finite element structural analysis of spacecraft