@article{997, abstract = {Recently it was shown that molecules rotating in superfluid helium can be described in terms of the angulon quasiparticles (Phys. Rev. Lett. 118, 095301 (2017)). Here we demonstrate that in the experimentally realized regime the angulon can be seen as a point charge on a 2-sphere interacting with a gauge field of a non-abelian magnetic monopole. Unlike in several other settings, the gauge fields of the angulon problem emerge in the real coordinate space, as opposed to the momentum space or some effective parameter space. Furthermore, we find a topological transition associated with making the monopole abelian, which takes place in the vicinity of the previously reported angulon instabilities. These results pave the way for studying topological phenomena in experiments on molecules trapped in superfluid helium nanodroplets, as well as on other realizations of orbital impurity problems.}, author = {Yakaboylu, Enderalp and Deuchert, Andreas and Lemeshko, Mikhail}, issn = {0031-9007}, journal = {Physical Review Letters}, number = {23}, publisher = {American Physical Society}, title = {{Emergence of non-abelian magnetic monopoles in a quantum impurity problem}}, doi = {10.1103/PhysRevLett.119.235301}, volume = {119}, year = {2017}, } @inproceedings{998, abstract = {A major open problem on the road to artificial intelligence is the development of incrementally learning systems that learn about more and more concepts over time from a stream of data. In this work, we introduce a new training strategy, iCaRL, that allows learning in such a class-incremental way: only the training data for a small number of classes has to be present at the same time and new classes can be added progressively. iCaRL learns strong classifiers and a data representation simultaneously. This distinguishes it from earlier works that were fundamentally limited to fixed data representations and therefore incompatible with deep learning architectures. We show by experiments on CIFAR-100 and ImageNet ILSVRC 2012 data that iCaRL can learn many classes incrementally over a long period of time where other strategies quickly fail. }, author = {Rebuffi, Sylvestre Alvise and Kolesnikov, Alexander and Sperl, Georg and Lampert, Christoph}, isbn = {978-153860457-1}, location = {Honolulu, HA, United States}, pages = {5533 -- 5542}, publisher = {IEEE}, title = {{iCaRL: Incremental classifier and representation learning}}, doi = {10.1109/CVPR.2017.587}, volume = {2017}, year = {2017}, } @inproceedings{999, abstract = {In multi-task learning, a learner is given a collection of prediction tasks and needs to solve all of them. In contrast to previous work, which required that annotated training data must be available for all tasks, we consider a new setting, in which for some tasks, potentially most of them, only unlabeled training data is provided. Consequently, to solve all tasks, information must be transferred between tasks with labels and tasks without labels. Focusing on an instance-based transfer method we analyze two variants of this setting: when the set of labeled tasks is fixed, and when it can be actively selected by the learner. We state and prove a generalization bound that covers both scenarios and derive from it an algorithm for making the choice of labeled tasks (in the active case) and for transferring information between the tasks in a principled way. We also illustrate the effectiveness of the algorithm on synthetic and real data. }, author = {Pentina, Anastasia and Lampert, Christoph}, isbn = {9781510855144}, location = {Sydney, Australia}, pages = {2807 -- 2816}, publisher = {ML Research Press}, title = {{Multi-task learning with labeled and unlabeled tasks}}, volume = {70}, year = {2017}, } @inproceedings{313, abstract = {Tunneling of a particle through a potential barrier remains one of the most remarkable quantum phenomena. Owing to advances in laser technology, electric fields comparable to those electrons experience in atoms are readily generated and open opportunities to dynamically investigate the process of electron tunneling through the potential barrier formed by the superposition of both laser and atomic fields. Attosecond-time and angstrom-space resolution of the strong laser-field technique allow to address fundamental questions related to tunneling, which are still open and debated: Which time is spent under the barrier and what momentum is picked up by the particle in the meantime? In this combined experimental and theoretical study we demonstrate that for strong-field ionization the leading quantum mechanical Wigner treatment for the time resolved description of tunneling is valid. We achieve a high sensitivity on the tunneling barrier and unambiguously isolate its effects by performing a differential study of two systems with almost identical tunneling geometry. Moreover, working with a low frequency laser, we essentially limit the non-adiabaticity of the process as a major source of uncertainty. The agreement between experiment and theory implies two substantial corrections with respect to the widely employed quasiclassical treatment: In addition to a non-vanishing longitudinal momentum along the laser field-direction we provide clear evidence for a non-zero tunneling time delay. This addresses also the fundamental question how the transition occurs from the tunnel barrier to free space classical evolution of the ejected electron.}, author = {Camus, Nicolas and Yakaboylu, Enderalp and Fechner, Lutz and Klaiber, Michael and Laux, Martin and Mi, Yonghao and Hatsagortsyan, Karen and Pfeifer, Thomas and Keitel, Cristoph and Moshammer, Robert}, issn = {1742-6588}, location = {Kazan, Russian Federation}, number = {1}, publisher = {American Physical Society}, title = {{Experimental evidence for Wigner's tunneling time}}, doi = {10.1088/1742-6596/999/1/012004}, volume = {999}, year = {2017}, } @article{391, abstract = {Three-dimensional topological insulators are bulk insulators with Z 2 topological electronic order that gives rise to conducting light-like surface states. These surface electrons are exceptionally resistant to localization by non-magnetic disorder, and have been adopted as the basis for a wide range of proposals to achieve new quasiparticle species and device functionality. Recent studies have yielded a surprise by showing that in spite of resisting localization, topological insulator surface electrons can be reshaped by defects into distinctive resonance states. Here we use numerical simulations and scanning tunnelling microscopy data to show that these resonance states have significance well beyond the localized regime usually associated with impurity bands. At native densities in the model Bi2X3 (X=Bi, Te) compounds, defect resonance states are predicted to generate a new quantum basis for an emergent electron gas that supports diffusive electrical transport. }, author = {Xu, Yishuai and Chiu, Janet and Miao, Lin and He, Haowei and Alpichshev, Zhanybek and Kapitulnik, Aharon and Biswas, Rudro and Wray, Lewis}, issn = {2041-1723}, journal = {Nature Communications}, publisher = {Springer Nature}, title = {{Disorder enabled band structure engineering of a topological insulator surface}}, doi = {10.1038/ncomms14081}, volume = {8}, year = {2017}, } @article{392, abstract = {We used femtosecond optical pump-probe spectroscopy to study the photoinduced change in reflectivity of thin films of the electron-doped cuprate La2-xCexCuO4 (LCCO) with dopings of x=0.08 (underdoped) and x=0.11 (optimally doped). Above Tc, we observe fluence-dependent relaxation rates that begin at a temperature similar to the one where transport measurements first show signatures of antiferromagnetic correlations. Upon suppressing superconductivity with a magnetic field, it is found that the fluence and temperature dependence of relaxation rates are consistent with bimolecular recombination of electrons and holes across a gap (2ΔAF) originating from antiferromagnetic correlations which comprise the pseudogap in electron-doped cuprates. This can be used to learn about coupling between electrons and high-energy (ω>2ΔAF) excitations in these compounds and set limits on the time scales on which antiferromagnetic correlations are static.}, author = {Vishik, Inna and Mahmood, Fahad and Alpichshev, Zhanybek and Gedik, Nuh and Higgins, Joshu and Greene, Richard}, issn = {2469-9969}, journal = {Physical Review B}, number = {11}, publisher = {American Physical Society}, title = {{Ultrafast dynamics in the presence of antiferromagnetic correlations in electron doped cuprate La2 xCexCuO4±δ}}, doi = {10.1103/PhysRevB.95.115125}, volume = {95}, year = {2017}, } @article{393, abstract = {We use a three-pulse ultrafast optical spectroscopy to study the relaxation processes in a frustrated Mott insulator Na2IrO3. By being able to independently produce the out-of-equilibrium bound states (excitons) of doublons and holons with the first pulse and suppress the underlying antiferromagnetic order with the second one, we were able to elucidate the relaxation mechanism of quasiparticles in this system. By observing the difference in the exciton dynamics in the magnetically ordered and disordered phases we found that the mass of this quasiparticle is mostly determined by its interaction with the surrounding spins. }, author = {Alpichshev, Zhanybek and Sie, Edbert and Mahmood, Fahad and Cao, Gang and Gedik, Nuh}, journal = {Physical Review B}, number = {23}, publisher = {American Physical Society}, title = {{Origin of the exciton mass in the frustrated Mott insulator Na2IrO3}}, doi = {10.1103/PhysRevB.96.235141}, volume = {96}, year = {2017}, } @inbook{424, abstract = {We show that very weak topological assumptions are enough to ensure the existence of a Helly-type theorem. More precisely, we show that for any non-negative integers b and d there exists an integer h(b, d) such that the following holds. If F is a finite family of subsets of Rd such that βi(∩G)≤b for any G⊊F and every 0 ≤ i ≤ [d/2]-1 then F has Helly number at most h(b, d). Here βi denotes the reduced Z2-Betti numbers (with singular homology). These topological conditions are sharp: not controlling any of these [d/2] first Betti numbers allow for families with unbounded Helly number. Our proofs combine homological non-embeddability results with a Ramsey-based approach to build, given an arbitrary simplicial complex K, some well-behaved chain map C*(K)→C*(Rd).}, author = {Goaoc, Xavier and Paták, Pavel and Patakova, Zuzana and Tancer, Martin and Wagner, Uli}, booktitle = {A Journey through Discrete Mathematics: A Tribute to Jiri Matousek}, editor = {Loebl, Martin and Nešetřil, Jaroslav and Thomas, Robin}, isbn = {978-331944479-6}, pages = {407 -- 447}, publisher = {Springer}, title = {{Bounding helly numbers via betti numbers}}, doi = {10.1007/978-3-319-44479-6_17}, year = {2017}, } @inproceedings{431, abstract = {Parallel implementations of stochastic gradient descent (SGD) have received significant research attention, thanks to its excellent scalability properties. A fundamental barrier when parallelizing SGD is the high bandwidth cost of communicating gradient updates between nodes; consequently, several lossy compresion heuristics have been proposed, by which nodes only communicate quantized gradients. Although effective in practice, these heuristics do not always converge. In this paper, we propose Quantized SGD (QSGD), a family of compression schemes with convergence guarantees and good practical performance. QSGD allows the user to smoothly trade off communication bandwidth and convergence time: nodes can adjust the number of bits sent per iteration, at the cost of possibly higher variance. We show that this trade-off is inherent, in the sense that improving it past some threshold would violate information-theoretic lower bounds. QSGD guarantees convergence for convex and non-convex objectives, under asynchrony, and can be extended to stochastic variance-reduced techniques. When applied to training deep neural networks for image classification and automated speech recognition, QSGD leads to significant reductions in end-to-end training time. For instance, on 16GPUs, we can train the ResNet-152 network to full accuracy on ImageNet 1.8 × faster than the full-precision variant. }, author = {Alistarh, Dan-Adrian and Grubic, Demjan and Li, Jerry and Tomioka, Ryota and Vojnović, Milan}, issn = {1049-5258}, location = {Long Beach, CA, United States}, pages = {1710--1721}, publisher = {Neural Information Processing Systems Foundation}, title = {{QSGD: Communication-efficient SGD via gradient quantization and encoding}}, volume = {2017}, year = {2017}, } @inproceedings{432, abstract = {Recently there has been significant interest in training machine-learning models at low precision: by reducing precision, one can reduce computation and communication by one order of magnitude. We examine training at reduced precision, both from a theoretical and practical perspective, and ask: is it possible to train models at end-to-end low precision with provable guarantees? Can this lead to consistent order-of-magnitude speedups? We mainly focus on linear models, and the answer is yes for linear models. We develop a simple framework called ZipML based on one simple but novel strategy called double sampling. Our ZipML framework is able to execute training at low precision with no bias, guaranteeing convergence, whereas naive quanti- zation would introduce significant bias. We val- idate our framework across a range of applica- tions, and show that it enables an FPGA proto- type that is up to 6.5 × faster than an implemen- tation using full 32-bit precision. We further de- velop a variance-optimal stochastic quantization strategy and show that it can make a significant difference in a variety of settings. When applied to linear models together with double sampling, we save up to another 1.7 × in data movement compared with uniform quantization. When training deep networks with quantized models, we achieve higher accuracy than the state-of-the- art XNOR-Net. }, author = {Zhang, Hantian and Li, Jerry and Kara, Kaan and Alistarh, Dan-Adrian and Liu, Ji and Zhang, Ce}, booktitle = {Proceedings of Machine Learning Research}, isbn = {978-151085514-4}, location = {Sydney, Australia}, pages = {4035 -- 4043}, publisher = {ML Research Press}, title = {{ZipML: Training linear models with end-to-end low precision, and a little bit of deep learning}}, volume = { 70}, year = {2017}, } @article{17616, abstract = {The optical and UV emission from sub-parsec massive black hole binaries (MBHBs) in active galactic nuclei (AGNs) is believed to vary periodically, on time-scales comparable to the binary's orbital time. If driven by accretion rate fluctuations, the variability could be isotropic. If dominated by relativistic Doppler modulation, the variability should instead be anisotropic, resembling a rotating forward-beamed lighthouse. We consider the infrared (IR) reverberation of either type of periodic emission by pc-scale circumbinary dust tori. We predict the phase and amplitude of IR variability as a function of the ratio of dust light crossing time to the source variability period, and of the torus inclination and opening angle. We enumerate several differences between the isotropic and anisotropic cases. Interestingly, for a nearly face-on binary with an inclined dust torus, the Doppler boost can produce IR variability without any observable optical/UV variability. Such orphan-IR variability would have been missed in optical searches for periodic AGNs. We apply our models to time-domain WISE IR data from the MBHB candidate PG 1302−102 and find consistency with dust reverberation by both isotropically emitting and Doppler-boosted sources in the shorter wavelength W1–W2 (2.8 → 5.3 μm) bands. We constrain the dust torus to be thin (aspect ratio ∼ 0.1), with an inner radius at 1–5 pc. More generally, our dust-echo models will aid in identifying new MBHB candidates, determining their nature and constraining the physical properties of MBHBs and their dust tori.}, author = {D'Orazio, Daniel J. and Haiman, Zoltán}, issn = {0035-8711}, journal = {Monthly Notices of the Royal Astronomical Society}, number = {1}, pages = {1198--1217}, publisher = {Oxford University Press}, title = {{Lighthouse in the dust: infrared echoes of periodic emission from massive black hole binaries★}}, doi = {10.1093/mnras/stx1269}, volume = {470}, year = {2017}, } @article{17630, abstract = {In symmetric gravitating systems experiencing rapid mass-loss, particle orbits change almost instantaneously, which can lead to the development of a sharply contoured density profile, including singular caustics for collisionless systems. This framework can be used to model a variety of dynamical systems, such as accretion discs following a massive black hole merger and dwarf galaxies following violent early star formation feedback. Particle interactions in the high-density peaks seem a promising source of observable signatures of these mass-loss events (i.e. a possible EM counterpart for black hole mergers or strong gamma-ray emission from dark matter annihilation around young galaxies), because the interaction rate depends on the square of the density. We study post-mass-loss density profiles, both analytic and numerical, in idealized cases and present arguments and methods to extend to any general system. An analytic derivation is presented for particles on Keplerian orbits responding to a drop in the central mass. We argue that this case, with initially circular orbits, gives the most sharply contoured profile possible. We find that despite the presence of a set of singular caustics, the total particle interaction rate is reduced compared to the unperturbed system; this is a result of the overall expansion of the system dominating over the steep caustics. Finally, we argue that this result holds more generally, and the loss of central mass decreases the particle interaction rate in any physical system.}, author = {Penoyre, Zephyr and Haiman, Zoltán}, issn = {0035-8711}, journal = {Monthly Notices of the Royal Astronomical Society}, number = {1}, pages = {498--512}, publisher = {Oxford University Press}, title = {{A drop in the pond: The effect of rapid mass-loss on the dynamics and interaction rate of collisionless particles}}, doi = {10.1093/mnras/stx2469}, volume = {473}, year = {2017}, } @article{17632, abstract = {We assess the contribution of dynamical hardening by direct three-body scattering interactions to the rate of stellar-mass black hole binary (BHB) mergers in galactic nuclei. We derive an analytic model for the single-binary encounter rate in a nucleus with spherical and disc components hosting a super-massive black hole (SMBH). We determine the total number of encounters NGW needed to harden a BHB to the point that inspiral due to gravitational wave emission occurs before the next three-body scattering event. This is done independently for both the spherical and disc components. Using a Monte Carlo approach, we refine our calculations for NGW to include gravitational wave emission between scattering events. For astrophysically plausible models, we find that typically NGW ≲ 10. We find two separate regimes for the efficient dynamical hardening of BHBs: (1) spherical star clusters with high central densities, low-velocity dispersions, and no significant Keplerian component and (2) migration traps in discs around SMBHs lacking any significant spherical stellar component in the vicinity of the migration trap, which is expected due to effective orbital inclination reduction of any spherical population by the disc. We also find a weak correlation between the ratio of the second-order velocity moment to velocity dispersion in galactic nuclei and the rate of BHB mergers, where this ratio is a proxy for the ratio between the rotation- and dispersion-supported components. Because discs enforce planar interactions that are efficient in hardening BHBs, particularly in migration traps, they have high merger rates that can contribute significantly to the rate of BHB mergers detected by the advanced Laser Interferometer Gravitational-Wave Observatory.}, author = {Leigh, N W C and Geller, A M and McKernan, B and Ford, K E S and Mac Low, M-M and Bellovary, J and Haiman, Zoltán and Lyra, W and Samsing, J and O'Dowd, M and Kocsis, B and Endlich, S}, issn = {0035-8711}, journal = {Monthly Notices of the Royal Astronomical Society}, number = {4}, pages = {5672--5683}, publisher = {Oxford University Press}, title = {{On the rate of black hole binary mergers in galactic nuclei due to dynamical hardening}}, doi = {10.1093/mnras/stx3134}, volume = {474}, year = {2017}, } @article{17633, abstract = {Using few-body simulations, we investigate the evolution of supermassive black holes (SMBHs) in galaxies (M* = 1010–1012 M⊙ at z = 0) at 0 < z < 4. Following galaxy merger trees from the Millennium simulation, we model BH mergers with two extreme binary decay scenarios for the ‘hard binary’ stage: a full or an empty loss cone. These two models should bracket the true evolution, and allow us to separately explore the role of dynamical friction and that of multibody BH interactions on BH mergers. Using the computed merger rates, we infer the stochastic gravitational wave background (GWB). Our dynamical approach is a first attempt to study the dynamical evolution of multiple SMBHs in the host galaxies undergoing mergers with various mass ratios (10−4 < q* < 1). Our main result demonstrates that SMBH binaries are able to merge in both scenarios. In the empty loss cone case, we find that BHs merge via multibody interactions, avoiding the ‘final parsec’ problem, and entering the pulsar timing arrays band with substantial orbital eccentricity. Our full loss cone treatment, albeit more approximate, suggests that the eccentricity becomes even higher when GWs become dominant, leading to rapid coalescences (binary lifetime ≲1 Gyr). Despite the lower merger rates in the empty loss cone case, due to their higher mass ratios and lower redshifts, the GWB in the full/empty loss cone models are comparable (0.70 × 10−15 and 0.53 × 10−15 at a frequency of 1 yr−1, respectively). Finally, we compute the effects of high eccentricities on the GWB spectrum.}, author = {Ryu, Taeho and Perna, Rosalba and Haiman, Zoltán and Ostriker, Jeremiah P. and Stone, Nicholas C.}, issn = {0035-8711}, journal = {Monthly Notices of the Royal Astronomical Society}, number = {3}, pages = {3410--3433}, publisher = {Oxford University Press}, title = {{Interactions between multiple supermassive black holes in galactic nuclei: A solution to the final parsec problem}}, doi = {10.1093/mnras/stx2524}, volume = {473}, year = {2017}, } @article{17667, abstract = {The Direct Collapse Black Hole (DCBH) scenario provides a solution for forming the massive black holes powering bright quasars observed in the early Universe. A prerequisite for forming a DCBH is that the formation of (much less massive) Population III stars be avoided - this can be achieved by destroying H2 via Lyman-Werner (LW) radiation (ELW = 12.6 eV). We find that two conditions must be met in the proto-galaxy that will host the DCBH. First, prior star formation must be delayed; this can be achieved with a background LW flux of JBG≳100 J21. Second, an intense burst of LW radiation from a neighbouring star-bursting proto-galaxy is required, just before the gas cloud undergoes gravitational collapse, to finally suppress star formation completely. We show here for the first time using high-resolution hydrodynamical simulations, including full radiative transfer, that this low-level background, combined with tight synchronisation and irradiation of a secondary proto-galaxy by a primary proto-galaxy, inevitably moves the secondary proto-galaxy onto the isothermal atomic cooling track, without the deleterious effects of either photo-evaporating the gas or polluting it by heavy elements. These, atomically cooled, massive proto-galaxies are expected to ultimately form a DCBH of mass 104−105M⊙.}, author = {Regan, John A. and Visbal, Eli and Wise, John H. and Haiman, Zoltán and Johansson, Peter H. and Bryan, Greg L.}, issn = {2397-3366}, journal = {Nature Astronomy}, number = {4}, publisher = {Springer Science and Business Media LLC}, title = {{Rapid formation of massive black holes in close proximity to embryonic protogalaxies}}, doi = {10.1038/s41550-017-0075}, volume = {1}, year = {2017}, } @article{17668, abstract = {Suppression of H2-cooling in early protogalaxies has important implications for the formation of supermassive black holes seeds, the first generation of stars, and the epoch of reionization. This suppression can occur via photodissociation of H2 (by ultraviolet Lyman-Werner [LW] photons) or by photodetachment of H-, a precursor in H2 formation (by infrared [IR] photons). Previous studies have typically adopted idealised spectra, with a blackbody or a power-law shape, in modeling the chemistry of metal-free protogalaxies, and utilised a single parameter, the critical UV flux, or Jcrit, to determine whether H2-cooling is prevented. This can be misleading, and that independent of the spectral shape, there is a critical curve in the (kLW , kH^- ) plane, where kLW and kH^- are the H2-dissocation rates by LW and IR photons, which determines whether a protogalaxy can cool below ~1000 Kelvin. We use a one-zone model to follow the chemical and thermal evolution of gravitationally collapsing protogalactic gas, to compute this critical curve, and provide an accurate analytical fit for it. We improve on previous works by considering a variety of more realistic Pop III or Pop II-type spectra from population synthesis models and perform fully frequency-dependent calculations of the H2-photodissociation rates for each spectrum. We compute the ratio kLW/kH^- for each spectrum, as well as the minimum stellar mass M_star, for various IMFs and metallicities, required to prevent cooling in a neighboring halo a distance d away. We provide critical M_star/d2 values for suppression of H2-cooling, with analytic fits, which can be used in future studies.}, author = {Wolcott-Green, J. and Haiman, Zoltán and Bryan, G. L.}, issn = {0035-8711}, journal = {Monthly Notices of the Royal Astronomical Society}, publisher = {Oxford University Press}, title = {{Beyond Jcrit: A critical curve for suppression of H2-cooling in protogalaxies}}, doi = {10.1093/mnras/stx167}, year = {2017}, } @article{17673, abstract = {Accurate forward modeling of weak lensing (WL) observables from cosmological parameters is necessary for upcoming galaxy surveys. Because WL probes structures in the non-linear regime, analytical forward modeling is very challenging, if not impossible. Numerical simulations of WL features rely on ray-tracing through the outputs of N-body simulations, which requires knowledge of the gravitational potential and accurate solvers for light ray trajectories. A less accurate procedure, based on the Born approximation, only requires knowledge of the density field, and can be implemented more efficiently and at a lower computational cost. In this work, we use simulations to show that deviations of the Born-approximated convergence power spectrum, skewness and kurtosis from their fully ray--traced counterparts are consistent with the smallest non-trivial O(Φ3) post-Born corrections (so-called geodesic and lens-lens terms). Our results imply a cancellation among the larger O(Φ4) (and higher order) terms, consistent with previous analytic work. We also find that cosmological parameter bias induced by the Born approximated power spectrum is negligible even for an LSST-like survey, once galaxy shape noise is considered. When considering higher order statistics such as the κ skewness and kurtosis, however, we find significant bias of up to 2.5σ. Using the LensTools software suite, we show that the Born approximation saves a factor of 4 in computing time with respect to the full ray-tracing in reconstructing the convergence.}, author = {Petri, Andrea and Haiman, Zoltán and May, Morgan}, issn = {2470-0010}, journal = {Physical Review D}, number = {12}, publisher = {American Physical Society}, title = {{Validity of the Born approximation for beyond Gaussian weak lensing observables}}, doi = {10.1103/physrevd.95.123503}, volume = {95}, year = {2017}, } @article{17696, abstract = {We utilize cosmological hydrodynamic simulations to study the formation of Population III (Pop III) stars in dark matter halos exposed to strong ionizing radiation. We simulate the formation of three halos subjected to a wide range of ionizing fluxes, and find that for high flux, ionization and photoheating can delay gas collapse and star formation up to halo masses significantly larger than the atomic cooling threshold. The threshold halo mass at which gas first collapses and cools increases with ionizing flux for intermediate values, and saturates at a value approximately an order of magnitude above the atomic cooling threshold for extremely high flux (e.g. ≈5×108 M⊙ at z≈6). This behavior can be understood in terms of photoheating, ionization/recombination, and Lyα cooling in the pressure-supported, self-shielded gas core at the center of the growing dark matter halo. We examine the spherically-averaged radial velocity profiles of collapsing gas and find that a gas mass of up to ≈106 M⊙ can reach the central regions within 3 Myr, providing an upper limit on the amount of massive Pop III stars that can form. The ionizing radiation increases this limit by a factor of a few compared to strong Lyman-Werner (LW) radiation alone. We conclude that the bright HeII 1640 Å emission recently observed from the high-redshift galaxy CR7 cannot be explained by Pop III stars alone. However, in some halos, a sufficient number of Pop III stars may form to be detectable with future telescopes such as the James Webb Space Telescope (JWST).}, author = {Visbal, Eli and Bryan, Greg L. and Haiman, Zoltán}, issn = {0035-8711}, journal = {Monthly Notices of the Royal Astronomical Society}, number = {2}, pages = {1456--1465}, publisher = {Oxford University Press}, title = {{What is the maximum mass of a Population III galaxy?}}, doi = {10.1093/mnras/stx909}, volume = {469}, year = {2017}, } @article{17698, abstract = {Gaseous circumbinary accretion discs provide a promising mechanism to facilitate the mergers of supermassive black holes (SMBHs) in galactic nuclei. We measure the torques exerted on accreting SMBH binaries, using 2D, isothermal, moving-mesh, viscous hydrodynamical simulations of circumbinary accretion discs. Our computational domain includes the entire inner region of the circumbinary disk with the individual black holes (BHs) included as point masses on the grid and a sink prescription to model accretion onto each BH. The BHs each acquire their own well-resolved accretion discs ("minidiscs"). We explore a range of mass removal rates for the sink prescription removing gas from the central regions of the minidiscs. We find that the torque exerted on the binary is primarily gravitational, and dominated by the gas orbiting close behind and ahead of the individual BHs. The torques from the distorted circumbinary disc farther out and from the direct accretion of angular momentum are subdominant. The torques are sensitive to the sink prescription: slower sinks result in more gas accumulating near the BHs and more negative torques, driving the binary to merger more rapidly. For faster sinks, the torques are less negative and eventually turn positive (for unphysically fast sinks). When the minidiscs are modeled as standard alpha discs, our results are insensitive to the choice of sink radius. Scaling the simulations to a binary orbital period tbin = 1yr and background disc accretion rate Mdot = 0.3MEdd in Eddington units, the binary inspirals on a timescale of 3X10^6 years, irrespective of the SMBH masses. For binaries with total mass <10^7Msun, this is shorter than the inspiral time due to gravitational wave (GW) emission alone, implying that gas discs will have a significant impact on the SMBH binary population and can affect the GW signal for Pulsar Timing Arrays.}, author = {Tang, Yike and MacFadyen, Andrew and Haiman, Zoltán}, issn = {0035-8711}, journal = {Monthly Notices of the Royal Astronomical Society}, number = {4}, pages = {4258--4267}, publisher = {Oxford University Press}, title = {{On the orbital evolution of supermassive black hole binaries with circumbinary accretion discs}}, doi = {10.1093/mnras/stx1130}, volume = {469}, year = {2017}, } @article{17706, abstract = {The Laser Interferometer Gravitational-Wave Observatory, LIGO, found direct evidence for double black hole binaries emitting gravitational waves. Galactic nuclei are expected to harbor the densest population of stellar-mass black holes. A significant fraction (∼30%) of these black holes can reside in binaries. We examine the fate of the black hole binaries in active galactic nuclei, which get trapped in the inner region of the accretion disk around the central supermassive black hole. We show that binary black holes can migrate into and then rapidly merge within the disk well within a Salpeter time. The binaries may also accrete a significant amount of gas from the disk, well above the Eddington rate. This could lead to detectable X-ray or gamma-ray emission, but would require hyper-Eddington accretion with a few percent radiative efficiency, comparable to thin disks. We discuss implications for gravitational wave observations and black hole population studies. We estimate that Advanced LIGO may detect ∼20 such, gas-induced binary mergers per year.}, author = {Bartos, Imre and Kocsis, Bence and Haiman, Zoltán and Márka, Szabolcs}, issn = {0004-637X}, journal = {The Astrophysical Journal}, number = {2}, publisher = {American Astronomical Society}, title = {{Rapid and bright stellar-mass binary black hole mergers in active galactic nuclei}}, doi = {10.3847/1538-4357/835/2/165}, volume = {835}, year = {2017}, }