Search

We report on the discovery of a very prominent mid-infrared (18-25 {\mu}m) excess associated with the trans-Neptunian dwarf planet (136472) Makemake. The excess, detected by the MIRI instrument of the James Webb Space Telescope, along with previous measurements from the Spitzer and Herschel space telescopes, indicates the occurrence of temperatures of about 150 K, much higher than what solid surfaces at Makemake's heliocentric distance could reach by solar irradiation. We identify two potential explanations: a continuously visible, currently active region, powered by subsurface upwelling and possibly cryovolcanic activity, covering <1% of Makemake's surface, or an as yet undetected ring containing very small carbonaceous dust grains, which have not been seen before in trans-Neptunian or Centaur rings. Both scenarios point to unprecedented phenomena among trans-Neptunian objects and could greatly impact our understanding of these distant worlds., Comment: Accepted for publication in The Astrophysical Journal Letters

Low-Rank Adaptation (LoRA) is a widely-used parameter-efficient finetuning method for large language models. LoRA saves memory by training only low rank perturbations to selected weight matrices. In this work, we compare the performance of LoRA and full finetuning on two target domains, programming and mathematics. We consider both the instruction finetuning (approximately 100K prompt-response pairs) and continued pretraining (20B unstructured tokens) data regimes. Our results show that, in the standard low-rank settings, LoRA substantially underperforms full finetuning. Nevertheless, LoRA better maintains the base model's performance on tasks outside the target domain. We show that LoRA mitigates forgetting more than common regularization techniques such as weight decay and dropout; it also helps maintain more diverse generations. Finally, we show that full finetuning learns perturbations with a rank that is 10-100X greater than typical LoRA configurations, possibly explaining some of the reported gaps. We conclude by proposing best practices for finetuning with LoRA., Comment: Final version with new experiments and analyses, as accepted to Transactions on Machine Learning Research, August 2024 (Featured Certification). https://openreview.net/forum?id=aloEru2qCG&noteId=Jb3PQNQDI2

Existing learning-based solutions to medical image segmentation have two important shortcomings. First, for most new segmentation task, a new model has to be trained or fine-tuned. This requires extensive resources and machine learning expertise, and is therefore often infeasible for medical researchers and clinicians. Second, most existing segmentation methods produce a single deterministic segmentation mask for a given image. In practice however, there is often considerable uncertainty about what constitutes the correct segmentation, and different expert annotators will often segment the same image differently. We tackle both of these problems with Tyche, a model that uses a context set to generate stochastic predictions for previously unseen tasks without the need to retrain. Tyche differs from other in-context segmentation methods in two important ways. (1) We introduce a novel convolution block architecture that enables interactions among predictions. (2) We introduce in-context test-time augmentation, a new mechanism to provide prediction stochasticity. When combined with appropriate model design and loss functions, Tyche can predict a set of plausible diverse segmentation candidates for new or unseen medical images and segmentation tasks without the need to retrain.

The NASA/DART (Double Asteroid Redirection Test) spacecraft successfully crashed on Dimorphos, the secondary component of the binary (65803) Didymos system. Following the impact, a large dust cloud was released, and a long-lasting dust tail was developed. We have extensively monitored the dust tail from the ground and from the Hubble Space Telescope (HST). We provide a characterization of the ejecta dust properties, i.e., particle size distribution and ejection speeds, ejection geometric parameters, and mass, by combining both observational data sets, and by using Monte Carlo models of the observed dust tail. The differential size distribution function that best fits the imaging data was a broken power-law, having a power index of --2.5 for particles of r$\le$ 3 mm, and of --3.7 for larger particles. The particles range in sizes from 1 $\mu$m up to 5 cm. The ejecta is characterized by two components, depending on velocity and ejection direction. The northern component of the double tail, observed since October 8th 2022, might be associated to a secondary ejection event from impacting debris on Didymos, although it is also possible that this feature results from the binary system dynamics alone. The lower limit to the total dust mass ejected is estimated at $\sim$6$\times$10$^6$ kg, half of this mass being ejected to interplanetary space., Comment: Accepted by Planetary Science Journal, July 7th, 2023

Hypernetworks, neural networks that predict the parameters of another neural network, are powerful models that have been successfully used in diverse applications from image generation to multi-task learning. Unfortunately, existing hypernetworks are often challenging to train. Training typically converges far more slowly than for non-hypernetwork models, and the rate of convergence can be very sensitive to hyperparameter choices. In this work, we identify a fundamental and previously unidentified problem that contributes to the challenge of training hypernetworks: a magnitude proportionality between the inputs and outputs of the hypernetwork. We demonstrate both analytically and empirically that this can lead to unstable optimization, thereby slowing down convergence, and sometimes even preventing any learning. We present a simple solution to this problem using a revised hypernetwork formulation that we call Magnitude Invariant Parametrizations (MIP). We demonstrate the proposed solution on several hypernetwork tasks, where it consistently stabilizes training and achieves faster convergence. Furthermore, we perform a comprehensive ablation study including choices of activation function, normalization strategies, input dimensionality, and hypernetwork architecture; and find that MIP improves training in all scenarios. We provide easy-to-use code that can turn existing networks into MIP-based hypernetworks., Comment: Source code at https://github.com/JJGO/hyperlight

While deep learning models have become the predominant method for medical image segmentation, they are typically not capable of generalizing to unseen segmentation tasks involving new anatomies, image modalities, or labels. Given a new segmentation task, researchers generally have to train or fine-tune models, which is time-consuming and poses a substantial barrier for clinical researchers, who often lack the resources and expertise to train neural networks. We present UniverSeg, a method for solving unseen medical segmentation tasks without additional training. Given a query image and example set of image-label pairs that define a new segmentation task, UniverSeg employs a new Cross-Block mechanism to produce accurate segmentation maps without the need for additional training. To achieve generalization to new tasks, we have gathered and standardized a collection of 53 open-access medical segmentation datasets with over 22,000 scans, which we refer to as MegaMedical. We used this collection to train UniverSeg on a diverse set of anatomies and imaging modalities. We demonstrate that UniverSeg substantially outperforms several related methods on unseen tasks, and thoroughly analyze and draw insights about important aspects of the proposed system. The UniverSeg source code and model weights are freely available at https://universeg.csail.mit.edu, Comment: Victor and Jose Javier contributed equally to this work. Project Website: https://universeg.csail.mit.edu

Convolutional Neural Networks (CNNs) are the predominant model used for a variety of medical image analysis tasks. At inference time, these models are computationally intensive, especially with volumetric data. In principle, it is possible to trade accuracy for computational efficiency by manipulating the rescaling factor in the downsample and upsample layers of CNN architectures. However, properly exploring the accuracy-efficiency trade-off is prohibitively expensive with existing models. To address this, we introduce Scale-Space HyperNetworks (SSHN), a method that learns a spectrum of CNNs with varying internal rescaling factors. A single SSHN characterizes an entire Pareto accuracy-efficiency curve of models that match, and occasionally surpass, the outcomes of training many separate networks with fixed rescaling factors. We demonstrate the proposed approach in several medical image analysis applications, comparing SSHN against strategies with both fixed and dynamic rescaling factors. We find that SSHN consistently provides a better accuracy-efficiency trade-off at a fraction of the training cost. Trained SSHNs enable the user to quickly choose a rescaling factor that appropriately balances accuracy and computational efficiency for their particular needs at inference., Comment: Code available at https://github.com/JJGO/scale-space-hypernetworks

Isostatic frames are mechanical networks that are simultaneously rigid and free of self-stress states, and is a powerful concept in understanding phase transitions in soft matter and designing of mechanical metamaterials. Here we analyze substructures of isostatic frames, by generalizing ``Assur graphs'' to the torus and examine them in physical systems. We show that the contact network of marginally jammed packings approach torus Assur graphs in the thermodynamic limit, and demostrate how Assur graphs offer a new design principle for mechanical metamaterials in which motion and stress can propagate in reconfigurable pathways, while rigidity of the entire structure is maintained., Comment: 8 pages, 5 figures

Linear complexity is an important parameter for arrays that are used in applications related to information security. In this work we survey constructions of two and three dimensional arrays, and present new results on the multidimensional linear complexity of periodic arrays obtained using the definition and method proposed in \cite{ArCaGoMoOrRuTi,GoHoMoRu,MoHoRu}. The results include a generalization of a bound for the linear complexity, a comparison with the measure of complexity for multisequences, and computations of the complexity of arrays with periods that are not relatively prime for which the ``unfolding method'' does not work. Conjectures for exact formulas and the asymptotic behavior of the complexity of some array constructions are formulated. We also present open source software for constructing multidimensional arrays and for computing their multidimensional linear complexity.

MIT wanted to commission a large scale artwork that would serve to 'illuminate a new campus gateway, inaugurate a space of exchange between MIT and Cambridge, and inspire our students, faculty, visitors, and the surrounding community to engage with art in new ways and to have art be part of their daily lives.' Among other things, the art was to reflect the fact that scientific discovery is often the result of many individual contributions, both acknowledged and unacknowledged. In this work, a group of computer scientists collaborated with a conceptual artist to produce a collective signature, or a signature learned from contributions of an entire community. After collecting signatures from two communities -- the university, and the surrounding city -- the computer scientists developed generative models and a human-in-the-loop feedback process to work with the artist create an original signature-like structure representative of each community. These signatures are now large-scale steel, LED and neon light sculptures that appear to sign two new buildings in Cambridge, MA.

MARCOT Pathfinder is a precursor for MARCOT (Multi Array of Combined Telescopes) at Calar Alto Observatory (CAHA) in Spain. MARCOT is intended to provide CARMENES, currently fiber-fed from the CAHA 3.5m Telescope, with a 5-15m light collecting area from a battery of several tens of small telescopes that are incoherently fed into the final joint single fiber feed of the spectrograph. The modular concept, based on commercially available telescopes, results in cost estimates that are a fraction of the ones for extremely large telescopes (ELT). As a novel approach, MARCOT will employ Multi-Mode Photonic Lanterns (MM-PL) that are being developed as a variant of classical photonic lanterns, to combine the light from the individual telescopes to a single fiber feed to the instrument. This progress report presents the overall concept of MARCOT, the pathfinder telescope and enclosure that is being commissioned at CAHA, the concept of MM-PL, and the next step of installing the Potsdam Multiplex Raman Spectrograph (MRS). MARCOT Pathfinder will be used to validate the conceptual design and predicted performance of MM-PL on sky with a 7 unit telescope prototype., Comment: 10 pages, 7 figures, submitted to SPIE

Observations that require large physical instrument dimensions and/or a considerable amount of cryogens, as it is for example the case for high spatial resolution far infrared astronomy, currently still face technological limits for their execution from space. The high cost and finality of space missions furthermore call for a very low risk approach and entail long development times. For certain spectral regions, prominently including the mid- to far-infrared as well as parts of the ultraviolet, stratospheric balloons offer a flexible and affordable complement to space telescopes, with short development times and comparatively good observing conditions. Yet, the entry burden to use balloon-borne telescopes is high, with research groups typically having to shoulder part of the infrastructure development as well. Aiming to ease access to balloon-based observations, we present the efforts towards a community-accessible balloon-based observatory, the European Stratospheric Balloon Observatory (ESBO). ESBO aims at complementing space-based and airborne capabilities over the next 10-15 years and at adding to the current landscape of scientific ballooning activities by providing a service-centered infrastructure for broader astronomical use, performing regular flights and offering an operations concept that provides researchers with a similar proposal-based access to observation time as practiced on ground-based observatories. We present details on the activities planned towards the goal of ESBO, the current status of the STUDIO (Stratospheric UV Demonstrator of an Imaging Observatory) prototype platform and mission, as well as selected technology developments with extensibility potential to space missions undertaken for STUDIO., Comment: Proc. 71st International Astronautical Congress, 12-14 October 2020. arXiv admin note: text overlap with arXiv:2012.14215, arXiv:2202.04580

Observations that require large physical instrument dimensions and/or a considerable amount of cryogens, as it is the case for high spatial resolution far infrared (FIR) astronomy, currently still face technological limits for their execution from space. Angular resolution and available observational capabilities are particularly affected. Balloon-based platforms promise to complement the existing observational capabilities by offering means to deploy comparatively large telescopes with comparatively little effort, including other advantages such as the possibility to regularly refill cryogens and to change and/or update instruments. The planned European Stratospheric Balloon Observatory (ESBO) aims at providing these additional large aperture FIR capabilities, exceeding the spatial resolution of Herschel, in the long term. The plans focus on reusable platforms performing regular flights and an operations concept that provides researchers with proposal-based access to observations. It thereby aims at offering a complement to other airborne, ground-based and space-based observatories in terms of access to wavelength regions, spatial resolution capability, and photometric stability. While the FIR capabilities are a main long-term objective, ESBO will offer benefits in other wavelength regimes along the way. Within the ESBO Design Study (ESBO DS), a prototype platform carrying a 0.5 m telescope for ultraviolet and visible light observations is being built and a platform concept for a next-generation FIR telescope is being studied. A flight of the UV/VIS prototype platform is estimated for 2021. In this paper we will outline the scientific and technical motivation for a large aperture balloon-based FIR observatory and the ESBO DS approach towards such an infrastructure. Secondly, we will present the technical motivation, science case, and instrumentation of the 0.5 m UV/VIS platform., Comment: Proc. 69th International Astronautical Congress, 1-5 October 2018. arXiv admin note: text overlap with arXiv:2111.11068

A stellar occultation by Pluto was observed on 6 June 2020 with the 1.3-m and 3.6-m telescopes located at Devasthal, Nainital, India, using imaging systems in the I and H bands, respectively. From this event, we derive a surface pressure for Pluto's atmosphere of $p_{\rm surf}= 12.23^{+0.65}_{-0.38} $~$\mu$bar. This shows that Pluto's atmosphere is in a plateau phase since mid-2015, a result which is in excellent agreement with the Pluto volatile transport model of Meza et al. (2019}. This value does not support the pressure decrease reported by independent teams, based on occultations observed in 2018 and 2019, see Young et al. (2021} and Arimatu et al. (2020), respectively., Comment: 8 pages, 3 figures, revised and accepted version for ApJL

This paper presents the concept of a community-accessible stratospheric balloon-based observatory that is currently under preparation by a consortium of European research institutes and industry. The planned European Stratospheric Balloon Observatory (ESBO) aims at complementing the current landscape of scientific ballooning activities by providing a service-centered infrastructure tailored towards broad astronomical use. In particular, the concept focuses on reusable platforms with exchangeable instruments and telescopes performing regular flights and an operations concept that provides researchers with options to test and operate own instruments, but later on also a proposal-based access to observations. It thereby aims at providing a complement to ground-, space-based, and airborne observatories in terms of access to wavelength regimes - particularly the ultraviolet (UV) and far infrared (FIR) regimes -, spatial resolution capability, and photometric stability. Within the currently ongoing ESBO Design Study (ESBO DS), financed within the European Union's Horizon 2020 Programme, a prototype platform carrying a 0.5-m telescope for UV and visible light observations is being built and concepts for larger following platforms, leading up to a next-generation FIR telescope are being studied. A flight of the UV/visible prototype platform is currently foreseen for 2021. We present the technical motivation, science case, instrumentation, and a two-stage image stabilization approach of the 0.5-m UV/visible platform. In addition, we briefly describe the novel mid-sized stabilized balloon gondola under design to carry telescopes in the 0.5 to 0.6 m range as well as the currently considered flight option for this platform. Secondly, we outline the scientific and technical motivation for a large balloon-based FIR telescope and the ESBO DS approach towards such an infrastructure., Comment: Proc. SPIE Conf., Astronomical Telescopes and Instrumentation, 10-15 June 2018

As datasets and models become increasingly large, distributed training has become a necessary component to allow deep neural networks to train in reasonable amounts of time. However, distributed training can have substantial communication overhead that hinders its scalability. One strategy for reducing this overhead is to perform multiple unsynchronized SGD steps independently on each worker between synchronization steps, a technique known as local SGD. We conduct a comprehensive empirical study of local SGD and related methods on a large-scale image classification task. We find that performing local SGD comes at a price: lower communication costs (and thereby faster training) are accompanied by lower accuracy. This finding is in contrast from the smaller-scale experiments in prior work, suggesting that local SGD encounters challenges at scale. We further show that incorporating the slow momentum framework of Wang et al. (2020) consistently improves accuracy without requiring additional communication, hinting at future directions for potentially escaping this trade-off.

We present results from the most comprehensive radio monitoring campaign towards the closest star to our Sun, Proxima Centauri. We report 1.1 to 3.1 GHz observations with the Australian Telescope Compact Array over 18 consecutive days in April 2017. We detect radio emission from Proxima Centauri for most of the observing sessions, which spanned $\sim$1.6 orbital periods of the planet Proxima b. The radio emission is stronger at the low-frequency band, centered around 1.6 GHz, and is consistent with the expected electron-cyclotron frequency for the known star's magnetic field intensity of about 600 Gauss. The 1.6 GHz light curve shows an emission pattern that is consistent with the orbital period of the planet Proxima b around the star Proxima, with its maxima of emission happening near the quadratures. We also observed two short-duration (a few minutes) flares and a long-duration (about three days) burst whose peaks happened close to the quadratures. We find that the frequency, large degree of circular polarization, change of the sign of circular polarization, and intensity of the observed radio emission are all consistent with expectations from electron cyclotron-maser emission arising from sub-Alfv\'enic star-planet interaction. We interpret our radio observations as signatures of interaction between the planet Proxima b and its host star Proxima. We advocate for monitoring other dwarf stars with planets to eventually reveal periodic radio emission due to star-planet interaction, thus opening a new avenue for exoplanet hunting and the study of a new field of exoplanet-star plasma interaction., Comment: 14 pages, 11 figures. Accepted on December 2, 2020 in A&A (main journal), several typos corrected