Search

We provide an original theoretical study of Inverse Reinforcement Learning (IRL) through the lens of reward compatibility, a novel framework to quantify the compatibility of a reward with the given expert's demonstrations. Intuitively, a reward is more compatible with the demonstrations the closer the performance of the expert's policy computed with that reward is to the optimal performance for that reward. This generalizes the notion of feasible reward set, the most common framework in the theoretical IRL literature, for which a reward is either compatible or not compatible. The grayscale introduced by the reward compatibility is the key to extend the realm of provably efficient IRL far beyond what is attainable with the feasible reward set: from tabular to large-scale MDPs. We analyze the IRL problem across various settings, including optimal and suboptimal expert's demonstrations and both online and offline data collection. For all of these dimensions, we provide a tractable algorithm and corresponding sample complexity analysis, as well as various insights on reward compatibility and how the framework can pave the way to yet more general problem settings.

In Reward Learning (ReL), we are given feedback on an unknown *target reward*, and the goal is to use this information to find it. When the feedback is not informative enough, the target reward is only *partially identifiable*, i.e., there exists a set of rewards (the feasible set) that are equally-compatible with the feedback. In this paper, we show that there exists a choice of reward, non-necessarily contained in the feasible set that, *depending on the ReL application*, improves the performance w.r.t. selecting the reward arbitrarily among the feasible ones. To this aim, we introduce a new *quantitative framework* to analyze ReL problems in a simple yet expressive way. We exemplify the framework in a *reward transfer* use case, for which we devise three provably-efficient ReL algorithms.

With the discovery of gravitational waves (GWs), the disks of Active Galactic Nuclei (AGN) have emerged as an interesting environment for hosting a fraction of their sources. AGN disks are conducive to forming both long and short Gamma-Ray Bursts (GRBs), and their anticipated cosmological occurrence within these disks has potential to serve as an independent tool for probing and calibrating the population of stars and compact objects within them, and their contribution to the GW-detected population. In this study, we employ Monte Carlo methods in conjunction with models for GRB electromagnetic emission in extremely dense media to simulate the cosmological occurrence of both long and short GRBs within AGN disks, while also estimating their detectability across a range of wavelengths, from gamma-rays to radio frequencies. {We investigate two extreme scenarios: ``undiffused", in which the radiation escapes without significant scattering (i.e. if the progenitor has excavated a funnel within the disk), and ``diffused", in which the radiation is propagated through the high-density medium, potentially scattered and absorbed. {In the diffused case,} we find that the majority of detectable GRBs are likely to originate from relatively low redshifts, and from the outermost regions of large supermassive black hole (SMBH) masses, $\gtrsim 10^{7.5} \rm M_{\odot}$. In the undiffused case, we expect a similar trend, but with a considerable contribution from the intermediate regions of lower SMBH masses. Detectable emission is generally expected to be dominant in prompt $\gamma$-rays if diffusion is not dominant, and X-ray afterglow if diffusion is important; however, the nature of the dominant observable signal highly depends on the specific AGN disk model, hence making GRBs in AGN disks also potential probes of the structure of the disks themselves., Comment: 18 pages, 16 figures

We present a novel implementation of a soft sphere, discrete elements code to simulate the dynamics of self-gravitating granular materials. The code is used to study the outcome of sub-sonic collisions between self-gravitating rubble piles with masses ranging from $\sim6\times10^{21}$ to $\sim6\times10^{22}$ g. These masses are representative of asteroids and planetesimals in the $\sim100$~km range. We simulate rubble piles composed of a range of particle sizes and analyze the collisions outcome focusing on the properties of the largest surviving fragment. We successfully test and validate the code against previous results. The results of our study show that rubble piles formed by collision of two parent rubble piles do not maintain the same particle size distribution as their parents. Rubble piles formed in low velocity collisions are characterized by a larger fraction of large particles, while the largest fragments of high-velocity collisions show a significant decrease in their mean particle size. We ascribe this effect to the fact that large particles transmit most of the forces during the collisions. In addition, we find that the mass of the largest post-collision fragment does depend on the rotation of the colliding rubble piles. This effect is especially noticeable when the pre-collision spin axes are parallel with each other and perpendicular to the relative velocity. This finding can be particularly relevant for meter to kilometer sized bodies embedded in protostellar accretion disks, where viscous stresses can efficiently align the target and projectile spin axes., Comment: 10 pages, 8 figures

Our goal is to extract useful knowledge from demonstrations of behavior in sequential decision-making problems. Although it is well-known that humans commonly engage in risk-sensitive behaviors in the presence of stochasticity, most Inverse Reinforcement Learning (IRL) models assume a risk-neutral agent. Beyond introducing model misspecification, these models do not directly capture the risk attitude of the observed agent, which can be crucial in many applications. In this paper, we propose a novel model of behavior in Markov Decision Processes (MDPs) that explicitly represents the agent's risk attitude through a utility function. We then define the Utility Learning (UL) problem as the task of inferring the observed agent's risk attitude, encoded via a utility function, from demonstrations in MDPs, and we analyze the partial identifiability of the agent's utility. Furthermore, we devise two provably efficient algorithms for UL in a finite-data regime, and we analyze their sample complexity. We conclude with proof-of-concept experiments that empirically validate both our model and our algorithms.

GRB 221009A was the Brightest gamma-ray burst Of All Time (BOAT), surpassing in prompt brightness all GRBs discovered in ~50 yr and in afterglow brightness in ~20 yr. We observed the BOAT with XMM-Newton 2.3 d after the prompt. The X-ray afterglow was still very bright and we collected the largest number of photons with the Reflection Grating Spectrometers (RGS) on a GRB. We searched the RGS data for narrow emission or absorption features. We did not detect any bright line feature. A candidate narrow feature is identified at a (rest-frame) energy of 1.455+0.006-0.014 keV, consistent with an Mg XII K{\alpha} emission line, slightly redshifted (0.012) with respect to the host galaxy. We assessed a marginal statistical significance of 3.0sigma for this faint feature based on conservative Monte Carlo simulations, which requires caution for any physical interpretation. If this line feature would be for real, we propose that it might originate from the reflection in the innermost regions of the infalling funnel from low-level late-time activity emission of the central engine., Comment: 8 pages, 3 figures. Accepted for publication in the Astrophysical Journal

We analyze the spectral evolution of 62 bright Fermi gamma-ray bursts with large enough signal to noise to allow for time resolved spectral analysis. We develop a new algorithm to test for single-pulse morphology that is insensitive to the specific shape of pulses. Instead, it only checks whether or not there are multiple, isolated, statistical significant peaks in the light curve. In addition, we carry out a citizen science test to assess light curve morphology and spectral evolution. We find that, no matter the adopted assessment method, bursts characterized by single-peaked prompt emission light curves have a greater tendency to also have a consistently decaying peak energy, or hard-to-soft spectral evolution. This contrasts the behavior of multi-peaked bursts, for which the tendency is to have a peak frequency that is not monotonically decreasing. We discuss this finding in the theoretical framework of internal/external shocks, and find it to be consistent with at least some single pulse bursts being associated with particularly high-density environments., Comment: 8 pages, 7 figures, Submitted to APJ

While Gamma Ray Bursts (GRBs) have the potential to shed light on the astrophysics of jets, compact objects, and cosmology, a major set back in their use as probes of these phenomena stems from our incomplete knowledge surrounding their prompt emission. There are numerous models that can account for various observations of GRBs in the gamma-ray and X-ray energy ranges due to the flexibility in the number of parameters that can be tuned to increase agreement with data. Furthermore, these models lack predictive power that can test future spectropolarimetric observations of GRBs across the electromagnetic spectrum. In this work, we use the MCRaT radiative transfer code to calculate the X-ray spectropolarimetric signatures expected from the photospheric model for two unique hydrodynamic simulations of long GRBs. We make time-resolved and time-integrated comparisons between the X-ray and gamma-ray mock observations, shedding light on the information that can be obtained from X-ray prompt emission signatures. Our results show that the $T_{90}$ derived from the X-ray lightcurve is the best diagnostic for the time that the central engine is active. We also find that our simulations reproduce the observed characteristics of the Einstein Probe detected GRB240315C. Based on our simulations, we are also able to make predictions for future X-ray spectropolarimetric measurements. Our results show the importance of conducting global radiative transfer calculations of GRB jets to better contextualize the prompt emission observations and constrain the mechanisms that produce the prompt emission., Comment: 14 pages, 9 figures with 2 movies, accepted to ApJ. Comments welcomed. Animations, datasets, and code to reproduce analysis can be found here: https://zenodo.org/doi/10.5281/zenodo.11623566

In online Inverse Reinforcement Learning (IRL), the learner can collect samples about the dynamics of the environment to improve its estimate of the reward function. Since IRL suffers from identifiability issues, many theoretical works on online IRL focus on estimating the entire set of rewards that explain the demonstrations, named the feasible reward set. However, none of the algorithms available in the literature can scale to problems with large state spaces. In this paper, we focus on the online IRL problem in Linear Markov Decision Processes (MDPs). We show that the structure offered by Linear MDPs is not sufficient for efficiently estimating the feasible set when the state space is large. As a consequence, we introduce the novel framework of rewards compatibility, which generalizes the notion of feasible set, and we develop CATY-IRL, a sample efficient algorithm whose complexity is independent of the cardinality of the state space in Linear MDPs. When restricted to the tabular setting, we demonstrate that CATY-IRL is minimax optimal up to logarithmic factors. As a by-product, we show that Reward-Free Exploration (RFE) enjoys the same worst-case rate, improving over the state-of-the-art lower bound. Finally, we devise a unifying framework for IRL and RFE that may be of independent interest., Comment: Advances in Neural Information Processing Systems 38 (NeurIPS 2024)

At least three members of the recently identified class of fast luminous blue optical transient show evidence of late-time electromagnetic activity in great excess of what predicted by an extrapolation of the early time emission. In particular, AT2022tsd displays fast, bright optical fluctuations approximately a month after the initial detection. Here, we propose that these transients are produced by exploding stars in black hole binary systems, and that the late-time activity is due to the accretion of clumpy ejecta onto the companion black hole. We derive the energetics and timescales involved, compute the emission spectrum, and discuss whether the ensuing emission is diffused or not in the remnant. We find that this model can explain the observed range of behaviors for reasonable ranges of the orbital separation and the ejecta velocity and clumpiness. Close separation and clumpy, high velocity ejecta result in bright variable emission, as seen in AT2022tsd. A wider separation and smaller ejecta velocity, conversely, give rise to fairly constant emission at a lower luminosity. We suggest that high-cadence, simultaneous, panchromatic monitoring of future transients should be carried out to better understand the origin of the late emission and the role of binarity in the diversity of explosive stellar transients., Comment: ApJL in press. Formally a revised version, but practically a completely new work that focuses on interpreting late activity in FLBOTs

The merger of two magnetized compact objects, such as neutron stars, forms a compact object which may launch a relativistic and collimated jet. Numerical simulations of the process show that a dense and highly magnetized medium surrounds the system. This study presents a semi-analytical model that models the effects that a static magnetized medium with a tangled field produces in relativistic, collimated, and non-magnetized jets. The model is a first approximation that addresses the magnetic field present in the medium and is based on pressure equilibrium principles between the jet, cocoon, and external medium. A fraction of the ambient medium field is allowed to be entrained in the cocoon. We find that the jet and cocoon properties may be affected by high magnetic fields ($\gtrsim 10^{15}$~G) and mixing. The evolution of the system may vary up to $\sim10\%$ (compared to the non-magnetized case). Low-mixing may produce a slower-broader jet with a broader and more energetic cocoon would be produced. On the other hand, high-mixing could produce a faster-narrower jet with a narrow and less-energetic cocoon. Two-dimensional hydrodynamical simulations are used to validate the model and to constrain the mixing parameter. Although the magnetic field and mixing have a limited effect, our semi-analytic model captures the general trend consistent with numerical results. For high magnetization, the results were found to be more consistent with the low mixing case in our semi-analytic model., Comment: 10 pages, 6 figures, 1 table, submitted to MNRAS

Long-duration gamma-ray bursts (LGRBs), thought to be produced during core-collapse supernovae, may have a prominent neutron component in the outflow material. If present, neutrons can change how photons scatter in the outflow by reducing its opacity, thereby allowing the photons to decouple sooner than if there were no neutrons present. Understanding the details of this process could therefore allow us to probe the central engine of LGRBs, which is otherwise hidden. Here, we present results of the photospheric emission from an LGRB jet, using a combination of relativistic hydrodynamic simulations and radiative transfer post-processing using the Monte Carlo Radiation Transfer (MCRaT) code. We control the size of the neutron component in the jet material by varying the equilibrium electron fraction $Y_{e}$, and we find that the presence of neutrons in the GRB fireball affects the Band parameters $\alpha$ and $E_{0}$, while the picture with the $\beta$ parameter is less clear. In particular, the break energy $E_{0}$ is shifted to higher energies. Additionally, we find that increasing the size of the neutron component also increases the total radiated energy of the outflow across multiple viewing angles. Our results not only shed light on LGRBs, but are also relevant to short-duration gamma-ray bursts associated with binary neutron star mergers, due to the likelihood of a prominent neutron component in such systems.

Inverse reinforcement learning (IRL) aims to recover the reward function of an expert agent from demonstrations of behavior. It is well-known that the IRL problem is fundamentally ill-posed, i.e., many reward functions can explain the demonstrations. For this reason, IRL has been recently reframed in terms of estimating the feasible reward set (Metelli et al., 2021), thus, postponing the selection of a single reward. However, so far, the available formulations and algorithmic solutions have been proposed and analyzed mainly for the online setting, where the learner can interact with the environment and query the expert at will. This is clearly unrealistic in most practical applications, where the availability of an offline dataset is a much more common scenario. In this paper, we introduce a novel notion of feasible reward set capturing the opportunities and limitations of the offline setting and we analyze the complexity of its estimation. This requires the introduction an original learning framework that copes with the intrinsic difficulty of the setting, for which the data coverage is not under control. Then, we propose two computationally and statistically efficient algorithms, IRLO and PIRLO, for addressing the problem. In particular, the latter adopts a specific form of pessimism to enforce the novel desirable property of inclusion monotonicity of the delivered feasible set. With this work, we aim to provide a panorama of the challenges of the offline IRL problem and how they can be fruitfully addressed., Comment: International Conference on Machine Learning 41 (ICML 2024)

We develop a general model of discrete choice that incorporates peer effects in preferences and consideration sets. We characterize the equilibrium behavior and establish conditions under which all parts of the model can be recovered from a sequence of choices. We allow peers to affect only preferences, only consideration, or both. We show that these peer-effect mechanisms have different behavioral implications in the data. This allows us to recover the set and the type of connections between the agents in the network. We then use this information to recover the preferences and the consideration mechanisms of each agent. These nonparametric identification results allow for general forms of heterogeneity across agents and do not rely on the variation of either exogenous covariates or the set of available options (menus). We apply our results to model expansion decisions by coffee chains and find evidence of limited consideration. We simulate counterfactual predictions and show how limited consideration slows down competition.

Inverse reinforcement learning (IRL) denotes a powerful family of algorithms for recovering a reward function justifying the behavior demonstrated by an expert agent. A well-known limitation of IRL is the ambiguity in the choice of the reward function, due to the existence of multiple rewards that explain the observed behavior. This limitation has been recently circumvented by formulating IRL as the problem of estimating the feasible reward set, i.e., the region of the rewards compatible with the expert's behavior. In this paper, we make a step towards closing the theory gap of IRL in the case of finite-horizon problems with a generative model. We start by formally introducing the problem of estimating the feasible reward set, the corresponding PAC requirement, and discussing the properties of particular classes of rewards. Then, we provide the first minimax lower bound on the sample complexity for the problem of estimating the feasible reward set of order ${\Omega}\Bigl( \frac{H^3SA}{\epsilon^2} \bigl( \log \bigl(\frac{1}{\delta}\bigl) + S \bigl)\Bigl)$, being $S$ and $A$ the number of states and actions respectively, $H$ the horizon, $\epsilon$ the desired accuracy, and $\delta$ the confidence. We analyze the sample complexity of a uniform sampling strategy (US-IRL), proving a matching upper bound up to logarithmic factors. Finally, we outline several open questions in IRL and propose future research directions., Comment: Submitted to: ICML23

Long and short gamma-ray bursts (GRBs), canonically separated at around 2 seconds duration, are associated with different progenitors: the collapse of a massive star and the merger of two compact objects, respectively. GRB 191019A was a long GRB ($T_{90}\sim64$ s). Despite the relatively small redshift z=0.248 and HST followup observations, an accompanying supernova was not detected. In addition, the host galaxy did not have significant star formation activity. Here we propose that GRB 191019A was produced by a binary compact merger, whose prompt emission was stretched in time by the interaction with a dense external medium. This would be expected if the burst progenitor was located in the disk of an active galactic nucleus, as supported by the burst localization close to the center of its host galaxy. We show that the light curve of GRB 191019A can be well modeled by a burst of intrinsic duration t=1.1 s and of energy $E_{\rm{iso}}=10^{51}$ erg seen moderately off-axis, exploding in a medium of density $10^7-10^8$ cm$^{-3}$. The double-peaked light curve carries the telltale features predicted for GRBs in high-density media, where the first peak is produced by the photosphere, and the second by the overlap of reverse shocks that take place before the internal shocks could happen. This would make GRB 191019A the first confirmed stellar explosion from within an accretion disk, with important implications for the formation and evolution of stars in accretion flows and for gravitational waves source populations., Comment: 6 pages, 2 figures. Accepted for publication in ApJL

Both long and short gamma-ray bursts (GRBs) are expected to occur in the dense environments of active galactic nuclei (AGN) accretion disks. As these bursts propagate through the disks they live in, they photoionize the medium causing time-dependent opacity that results in transients with unique spectral evolution. In this paper we use a line-of-sight radiation transfer code coupling metal and dust evolution to simulate the time-dependent absorption that occurs in the case of both long and short GRBs. Through these simulations, we investigate the parameter space in which dense environments leave a potentially observable imprint on the bursts. Our numerical investigation reveals that time dependent spectral evolution is expected for central supermassive black hole masses between $10^5$ and $5 \times 10^7$ solar masses in the case of long GRBs, and between $10^4$ and $10^7$ solar masses in the case of short GRBs. Our findings can lead to the identification of bursts exploding in AGN disk environments through their unique spectral evolution coupled with a central location. In addition, the study of the time-dependent evolution would allow for studying the disk structure, once the identification with an AGN has been established. Finally, our findings lead to insight into whether GRBs contribute to the AGN emission, and which kind, thus helping to answer the question of whether GRBs can be the cause of some of the as-of-yet unexplained AGN time variability.

The merger of two neutron stars (NSs) produces the emission of gravitational waves, the formation of a compact object surrounded by a dense and magnetized environment, and the launching of a collimated and relativistic jet, which will eventually produce a short gamma-ray burst (SGRB). The interaction of the jet with the environment has been shown to play a major role in shaping the structure of the outflow that eventually powers the gamma-ray emission. In this paper, we present a set of 2.5 dimensional RMHD simulations that follow the evolution of a relativistic non-magnetized jet through a medium with different magnetization levels, as produced after the merger of two NSs. We find that the predominant consequence of a magnetized ambient medium is that of suppressing instabilities within the jet, and preventing the formation of a series of collimation shocks. One implication of this is that internal shocks lose efficiency, causing bursts with low-luminosity prompt emission. On the other hand, the jet-head velocity and the induced magnetization within the jet are fairly independent from the magnetization of the ambient medium. Future numerical studies with a larger domain are necessary to obtain light curves and spectra in order to better understand the role of magnetized media., Comment: 7 pages, 5 figures, 1 table. Submitted to MNRAS

Long and short gamma-ray bursts are traditionally associated with galactic environments, where circumburst densities are small or moderate (few to hundreds of protons per cubic cm). However, both are also expected to occur in the disks of Active Galactic Nuclei, where the ambient medium density can be much larger. In this work we study, via semi-analytical methods, the propagation of the GRB outflow, its interaction with the external material, and the ensuing prompt radiation. In particular, we focus on the case in which the external shock develops early in the evolution, at a radius that is smaller than the internal shock one. We find that bursts in such high density environments are likely characterized by a single, long emission episode that is due to the superposition of individual pulses, with a characteristic hard to soft evolution irrespective of the light curve luminosity. While multi-pulse light curves are not impossible, they would require the central engine to go dormant for a long time before re-igniting. In addition, short GRB engines would produce bursts with prompt duration that would exceed the canonical 2 s separation threshold and would likely be incorrectly classified as long events, even though they would not be accompanied by a simultaneous supernova. Finally, these events have a large dynamical efficiency which would produce a bright prompt emission followed by a somewhat dim afterglow., Comment: 9 pages, 5 figures, under review for AAS Journals

The origin of short gamma-ray bursts (sGRBs) is associated with outflows powered by the remnant of a binary neutron star merger. This remnant can be either a black hole or a highly magnetized, fastly spinning neutron star, also known as a magnetar. Here, we present the results of two relativistic magnetohydrodynamical (RMHD) simulations aimed at investigating the large-scale dynamics and propagation of magnetar collimated outflows through the medium surrounding the remnant. The first simulation evolves a realistic jet by injecting external simulation data, while the second evolves an analytical model jet with similar properties for comparison. We find that both outflows remain collimated and successfully emerge through the static medium surrounding the remnant. However, they fail to attain relativistic velocities and only reach a mean maximum speed of ~0.7c for the realistic jet, and ~0.6c for the analytical jet. We also find that the realistic jet has a much more complex structure. The lack of highly relativistic speeds, that makes these jets unsuitable as short GRB sources, is due to numerical limitations and not general to all possible magnetar outflows. A jet like the one we study, however, could give rise to or augment a blue kilonova component. In addition, it would make the propagation of a relativistic jet easier, should one be launched after the neutron star collapses into a black hole., Comment: 11 pages, 7 figures, accepted to ApJ

The disks of Active Galactic Nuclei (AGNs) have emerged as rich environments for the production and capture of stars and the compact objects that they leave behind. These stars produce long Gamma-Ray Bursts (LGRBs) at their deaths, while frequent interactions among compact objects form binary neutron stars and neutron star-black hole binaries, leading to short GRBs (SGRBs) upon their merger. Predicting the properties of these transients as they emerge from the dense environments of AGN disks is key to their proper identification and to better constrain the star and compact object population in AGN disks. Some of these transients would appear unusual because they take place in much higher densities than the interstellar medium. Others, which are the subject of this paper, would additionally be modified by radiation diffusion since they are generated within optically thick regions of the accretion disks. Here we compute the GRB afterglow light curves for diffused GRB sources for a representative variety of central black-hole masses and disk locations. We find that the radiation from radio to UV and soft X-rays can be strongly suppressed by synchrotron self-absorption in the dense medium of the AGN disk. In addition, photon diffusion can significantly delay the emergence of the emission peak, turning a beamed, fast transient into a slow, isotropic, and dimmer one. These would appear as broadband-correlated AGN variability with dominance at the higher frequencies. Their properties can constrain both the stellar populations within AGN disks as well as the disk structure., Comment: Accepted for publication in MNRAS

GW170817 is the first binary neutron star (NS) merger detected in gravitational waves (GWs) and photons, and so far remains the only GW event of its class with a definitive electromagnetic (EM) counterpart. Radio emission from the structured jet associated with GW170817 has faded below the sensitivity achievable via deep radio observations with the most sensitive radio arrays currently in operation. Hence, we now have the opportunity to probe the radio re-brightening that some models predict, should emerge at late times from the interaction of the dynamically-stripped merger ejecta with the interstellar medium. Here we present the latest results from our deep radio observations of the GW170817 field with the Karl G. Jansky Very Large Array (VLA), 4.5 years after the merger. Our new data at $3\,$GHz do not show any compelling evidence for emission in excess to the tail of the jet afterglow ($<3.3\,\mu$Jy), confirming our previous results. We thus set new constraints on the dynamical ejecta afterglow models. These constraints favor single-speed ejecta with energy $\lesssim 10^{50}\,$erg (for an ejecta speed of $\beta_0=0.5$), or steeper energy-speed distributions of the kilonova ejecta. Our results also suggest larger values of the cold, non-rotating maximum NS mass in equal mass scenarios. However, without a detection of the dynamical ejecta afterglow, obtaining precise constraints on the NS equation of state remains challenging., Comment: 6 pages, 5 figures. Published in ApJ

The detection of electromagnetic radiation (EM) accompanying the gravitational wave (GW) signal from the binary neutron star (BNS) merger GW170817 has revealed that these systems constitute at least a fraction of the progenitors of short gamma-ray bursts (SGRBs). As gravitational wave detectors keep pushing their detection horizons, it is important to assess coupled GW/EM probabilities, and how to maximize observational prospects. Here we perform population synthesis calculations of BNS evolution with the code MOBSE, and seed the binaries in galaxies at three representative redshifts (z=0.01,0.1,1) of the Illustris TNG50 simulation. The binaries are evolved and their locations numerically tracked in the host galactic potentials until merger. Adopting the astrophysical parameters of GRB170817A as a prototype, we numerically compute the broadband lightcurves of jets from BNS mergers, with the afterglow brightness depending on the local medium density at the merger sites. We perform Monte Carlo simulations of the resulting EM population assuming either a random viewing angle with respect to the jet, or a jet aligned with the orbital angular momentum of the binary, which biases the viewing angle probability for GW-triggered events. We find that ~70-80% of BNSs from z=0.01 should be detectable in gamma-rays. The afterglow detection probabilities of GW-triggered BNS mergers vary between ~0.3-0.7%, with higher values for jets aligned with the BNS angular momentum, and are comparable across the high and low-energy bands, unlike gamma-ray-triggered events (cosmological SGRBs) which are significantly brighter at higher energies. We further quantify observational biases with respect to host galaxy masses., Comment: 15 pages, 12 figures, submitted to MNRAS

The study of Gamma Ray Bursts (GRBs) has the potential to improve our understanding of high energy astrophysical phenomena. In order to reliably use GRBs to this end, we first need to have a well-developed grasp of the mechanism that produces the radiation within GRB jets and how that relates to their structure. One model for the emission mechanism of GRBs invokes radiation produced deep in the jet which eventually escapes the jet at its photosphere. While this model has been able to explain a number of observed GRB characteristics, it is currently lacking in predictive power and in ability to fully reproduce GRB spectra. In order to address these shortcomings of the model, we have expanded the capabilities of the MCRaT code, a state of the art radiative transfer code that can now simulate optical to gamma ray radiation propagating in a hydrodynamically simulated GRB jet. Using the MCRaT code, we have constructed mock observed light curves, spectra, and polarization from optical to gamma ray energies for the simulated GRBs. Using these mock observables, we have compared our simulations of photospheric emission to observations and found much agreement between the two. Furthermore, the MCRaT calculations combined with the hydrodynamical simulations allow us to connect the mock observables to the structure of the simulated GRB jet in a way that was not previously possible. While there are a number of improvements that can be made to the analyses, the steps taken here begin to pave the way for us to fully understand the connection between the structure of a given GRB jet and the radiation that would be expected from it., Comment: 10 pages, 3 figures, accepted as a conference proceeding to the Sixteenth Marcel Grossmann Meeting

Although Gamma Ray Bursts (GRBs) have been detected for many decades, the lack of knowledge regarding the radiation mechanism that produces the energetic flash of radiation, or prompt emission, from these events has prevented the full use of GRBs as probes of high energy astrophysical processes. While there are multiple models that attempt to describe the prompt emission, each model can be tuned to account for observed GRB characteristics in the gamma and X-ray energy bands. One energy range that has not been fully explored for the purpose of prompt emission model comparison is that of the optical band, especially with regards to polarization. Here, we use an improved MCRaT code to calculate the expected photospheric optical and gamma-ray polarization signatures ($\Pi_\mathrm{opt}$ and $\Pi_\gamma$, respectively) from a set of two relativistic hydrodynamic long GRB simulations, which emulate a constant and variable jet. We find that time resolved $\Pi_\mathrm{opt}$ can be large ($\sim 75\%$) while time-integrated $\Pi_\mathrm{opt}$ can be smaller due to integration over the asymmetries in the GRB jet where optical photons originate; $\Pi_\gamma$ follows a similar evolution as $\Pi_\mathrm{opt}$ with smaller polarization degrees. We also show that $\Pi_\mathrm{opt}$ and $\Pi_\gamma$ agree well with observations in each energy range. Additionally, we make predictions for the expected polarization of GRBs based on their location within the Yonetoku relationship. While improvements can be made to our analyses and predictions, they exhibit the insight that global radiative transfer simulations of GRB jets can provide with respect to current and future observations., Comment: 16 pages; 7 figures including 2 animations (animations can be obtained from: https://github.com/parsotat/SpectropolarimetryAnimations); accepted to ApJ; the MCRaT and ProcessMCRaT codes are open source and available at: github.com/lazzati-astro/MCRaT and at: github.com/parsotat/ProcessMCRaT

The association of GRB170817A with GW170817 has confirmed the long-standing hypothesis that binary neutron star (BNS) mergers are the progenitors of at least some short gamma-ray bursts (SGRBs). This connection has ushered in an era in which broadband observations of SGRBs, together with measurements of the time delay between the gravitational waves and the electromagnetic radiation, allow to probe the properties of the emitting outflow and its engine to an unprecedented detail. Since the structure of the radiating outflow is molded by the interaction of a relativistic jet with the binary ejecta, it is of paramount importance to study the system in a realistic setting. Here we present a three-dimensional hydrodynamic simulation of a relativistic jet propagating in the ejecta of a BNS merger, which were computed with a general relativistic magnetohydrodynamic simulation. We find that the jet's centroid oscillates around the axis of the system, due to inhomogeneities encountered in the propagation. These oscillations allow the jet to find the path of least resistance and travel faster than an identical jet in smooth ejecta. In our setup the breakout time is ~0.6 sec, comparable to the expected central engine duration in SGRBs and possibly a non-negligible fraction of the total delay between the gravitational and gamma-ray signals. Our simulation also shows that energy is carried in roughly equal amounts by the jet and by the cocoon, and that about 20 per cent of the injected energy is transferred to the ejecta via mechanical work., Comment: Under review for publication in ApJL

A complete understanding of Gamma Ray Bursts (GRBs) has been difficult to achieve due to our incomplete knowledge of the radiation mechanism that is responsible for producing the prompt emission. This emission, which is detected in the first tens of seconds of the GRB, is typically dominated by hard X-ray and gamma ray photons although, there have also been a few dozen prompt optical detections. These optical detections have the potential to discriminate between plausible prompt emission models, such as the photospheric and synchrotron shock models. In this work we use an improved MCRaT code, which includes cyclo-synchrotron emission and absorption, to conduct radiative transfer calculations from optical to gamma ray energies under the photospheric model. The calculations are conducted using a set of two dimensional relativistic hydrodynamic long GRB jet simulations, consisting of a constant and variable jet. We predict the correlations between the optical and gamma ray light curves as functions of observer angle and jet variability and find that there should be extremely dim optical prompt precursors for large viewing angles. Additionally, the detected optical emission originates from dense regions of the outflow such as shock interfaces and the jet-cocoon interface. Our results also show that the photospheric model is not able to account for the current set of optical prompt detections that have been made and additional radiative mechanisms are needed to explain these prompt optical observations. These findings show the importance of conducting global radiative transfer simulations using hydrodynamically calculated jet structures., Comment: 22 pages, with 5 pages for the appendix; 11 figures; accepted to ApJ; the MCRaT code is open source and available at: https://github.com/lazzati-astro/MCRaT and the ProcessMCRaT code is also open source and available at: https://github.com/parsotat/ProcessMCRaT

GRB 190114C was a bright burst that occurred in the local Universe (z=0.425). It was the first gamma-ray burst (GRB) ever detected at TeV energies, thanks to MAGIC. We characterize the ambient medium properties of the host galaxy through the study of the absorbing X-ray column density. Joining Swift, XMM-Newton, and NuSTAR observations, we find that the GRB X-ray spectrum is characterized by a high column density that is well in excess of the expected Milky Way value and decreases, by a factor of ~2, around ~$10^5$ s. Such a variability is not common in GRBs. The most straightforward interpretation of the variability in terms of photoionization of the ambient medium is not able to account for the decrease at such late times, when the source flux is less intense. Instead, we interpret the decrease as due to a clumped absorber, denser along the line of sight and surrounded by lower-density gas. After the detection at TeV energies of GRB 190114C, two other GRBs were promptly detected. They share a high value of the intrinsic column density and there are hints for a decrease of the column density, too. We speculate that a high local column density might be a common ingredient for TeV-detected GRBs., Comment: 6 pages, 4 figures. Accepted for publication in A&A

We present new radio observations of the binary neutron star merger GW170817 carried out with the Karl G. Jansky Very large Array (VLA) more than 3\,yrs after the merger. Our combined dataset is derived by co-adding more than $\approx32$\,hours of VLA time on-source, and as such provides the deepest combined observation (rms sensitivity $\approx 0.99\,\mu$Jy) of the GW170817 field obtained to date at 3\,GHz. We find no evidence for a late-time radio re-brightening at a mean epoch of $t\approx 1200$\,d since merger, in contrast to a $\approx 2.1\,\sigma$ excess observed at X-ray wavelengths at the same mean epoch. Our measurements agree with expectations from the post-peak decay of the radio afterglow of the GW170817 structured jet. Using these results, we constrain the parameter space of models that predict a late-time radio re-brightening possibly arising from the high-velocity tail of the GW170817 kilonova ejecta, which would dominate the radio and X-ray emission years after the merger (once the structured jet afterglow fades below detection level). Our results point to a steep energy-speed distribution of the kilonova ejecta (with energy-velocity power law index $\alpha \gtrsim 5$). We suggest possible implications of our radio analysis, when combined with the recent tentative evidence for a late-time re-brightening in the X-rays, and highlight the need for continued radio-to-X-ray monitoring to test different scenarios., Comment: 6 pages, 2 figures. Version published in ApJL (includes minor updates to figures, data and refs)

The disks of Active Galactic Nuclei (AGNs), traditionally studied as the feeders of the supermassive black holes (SMBHs) at their centers, have recently triggered a lot of interest also as hosts to massive stars and hence their neutron star (NS) and black hole (BH) remnants. Migration traps and gas torques in these disks favor binary formation and enhance the rate of compact object mergers. In these environments both long gamma-ray bursts (GRBs) from the death of massive stars and short GRBs from NS-NS and NS-BH mergers are expected. However, their properties in the unique environments of AGN disks have never been studied. Here we show that GRBs in AGNs can display unique features, owing to the unusual relative position of the shocks that characterize the burst evolution and the Thomson photosphere of the AGN disk. In dense environments, for example, the external shock develops before the internal shocks, leading to prompt emission powered by a relativistic reverse shock. The transient's time evolution is also compressed, yielding afterglow emission that is much brighter and peaks much earlier than for GRBs in the interstellar medium. Additionally, in regions of the disk that are sufficiently dense and extended, the light curves are dominated by diffusion, since the fireball is trapped inside the disc photosphere. These diffusion-dominated transients emerge on timescales of days in disks around SMBHs of $\sim 10^6 M_\odot$ to years for SMBHs of $\sim 10^8 M_\odot$. Finally, a large fraction of events, especially in AGNs with SMBHs $\lesssim 10^7 M_\odot$, display time-variable absorption in the X-ray band., Comment: Submitted to ApJ Letters

The detection of GW170817, it's extensive multi-wavelength follow-up campaign, and the large amount of theoretical development and interpretation that followed, have resulted in a significant step forward in the understanding of the binary neutron star merger phenomenon as a whole. One of its aspects is seeing the merger as a progenitor of short gamma-ray bursts (SGRB), which will be the subject of this review. On the one hand, GW170817 observations have confirmed some theoretical expectations, exemplified by the confirmation that binary neutron star mergers are the progenitors of SGRBs. In addition, the multimessenger nature of GW170817 has allowed for gathering of unprecedented data, such as the trigger time of the merger, the delay with which the gamma-ray photons were detected, and the brightening afterglow of an off-axis event. All together, the incomparable richness of the data from GW170817 has allowed us to paint a fairly detailed picture of at least one SGRB. I will detail what we learned, what new questions have arisen, and the perspectives for answering them when a sample of GW170817-comparable events have been studied., Comment: Mini-review for Frontiers. Final published version

We present the full panchromatic afterglow light curve data of GW170817, including new radio data as well as archival optical and X-ray data, between 0.5 and 940 days post-merger. By compiling all archival data, and reprocessing a subset of it, we have evaluated the impact of differences in data processing or flux determination methods used by different groups, and attempted to mitigate these differences to provide a more uniform dataset. Simple power-law fits to the uniform afterglow light curve indicate a $t^{0.86\pm0.04}$ rise, a $t^{-1.92\pm0.12}$ decline, and a peak occurring at $155\pm4$ days. The afterglow is optically thin throughout its evolution, consistent with a single spectral index ($-0.584\pm0.002$) across all epochs. This gives a precise and updated estimate of the electron power-law index, $p=2.168\pm0.004$. By studying the diffuse X-ray emission from the host galaxy, we place a conservative upper limit on the hot ionized ISM density, $<$0.01 cm$^{-3}$, consistent with previous afterglow studies. Using the late-time afterglow data we rule out any long-lived neutron star remnant having magnetic field strength between 10$^{10.4}$ G and 10$^{16}$ G. Our fits to the afterglow data using an analytical model that includes VLBI proper motion from Mooley et al. (2018), and a structured jet model that ignores the proper motion, indicates that the proper motion measurement needs to be considered while seeking an accurate estimate of the viewing angle., Comment: Accepted by ApJ. Revised X-ray and up-to-date afterglow measurements are available in ASCII format on GitHub (https://github.com/kmooley/GW170817/) and mirrored on tauceti.caltech.edu (http://www.tauceti.caltech.edu/kunal/gw170817/). The online dataset will be continuously updated as new measurements become available

A comprehensive understanding of Gamma Ray Bursts (GRBs) has been elusive due to the variety of questions surrounding the radiation mechanism at play in these events. Polarization measurements of GRBs can heavily constrain the relevant radiation mechanisms and the structure of the GRB jet; however, there is a limited number of theoretical predictions that observed GRB polarizations can be compared against. Here, we conduct radiative transfer calculations of a set of two dimensional relativistic hydrodynamic long GRB (LGRB) jet simulations, of a constant and a variable jet, using the Monte Carlo Radiation Transport (MCRaT) code. MCRaT has been enhanced by the inclusion of polarization; it has been first verified by reproducing a variety of results in the literature and then used to obtain the time integrated and resolved polarization degrees and angles of the synthetic LGRBs. While the obtained time-integrated polarization degrees {($\lesssim 1$\%)} are consistent with the constraints from the POLAR experiment, they are lower than other theoretical studies due to the lack of strong gradients in the model jet profiles that we use. The time resolved results suggests that GRBs with wide jets observed on axis will have small polarization degrees ($\lesssim 2\%$) and {constant polarization angles}, during the brightest portion of the light curve. GRBs observed off axis will have larger polarization degrees and polarization angles that change with the temporal structure of radiating shells in the outflow. We then place our results in the context of GRB prompt emission models and future LEAP and POLAR-2 GRB polarimetry detections., Comment: 14 pages, 7 figures, final version after it has been accepted to ApJ

GRB 170817A was a subluminous short gamma-ray burst detected about 1.74 s after the gravitational wave signal GW170817 from a binary neutron star (BNS) merger. It is now understood as an off-axis event powered by the cocoon of a relativistic jet pointing 15 to 30 degrees away from the direction of observation. The cocoon was energized by the interaction of the incipient jet with the non-relativistic baryon wind from the merger remnant, resulting in a structured outflow with a narrow core and broad wings. In this paper, we couple the observational constraints on the structured outflow with a model for the jet-wind interaction to constrain the intrinsic properties with which the jet was launched by the central engine, including its time delay from the merger event. Using wind prescriptions inspired by magnetized BNS merger simulations, we find that the jet was launched within about 0.4 s from the merger, implying that the 1.74 s observed delay was dominated by the fireball propagation up to the photospheric radius. We also constrain, for the first time for any gamma-ray burst, the jet opening angle at injection and set a lower limit to its asymptotic Lorentz factor. These findings suggest an initially Poynting-flux dominated jet, launched via electromagnetic processes. If the jet was powered by an accreting black hole, they also provide a significant constraint on the survival time of the metastable neutron star remnant., Comment: Accepted for publication in ApJ