搜索结果：dawn

A plausible scene evolution depends on the maneuver being considered, while a good maneuver depends on how the scene may evolve. Existing World Action Models (WAMs) largely miss this reciprocity, treating world prediction and action generation as either isolated parallel branches or rigid predict-then-plan pipelines. We formalize this perspective as World-Action Interactive Models (WAIMs), and instantiate it in autonomous driving with \textbf{DAWN} (\textbf{D}enoising \textbf{A}ctions and \textbf{W}orld i\textbf{N}teractive model), a simple yet strong latent generative baseline. DAWN operates in a compact semantic latent space and couples a \emph{World Predictor} with a \emph{World-Conditioned Action Denoiser}: the predicted world hypothesis conditions action denoising, while the denoised action hypothesis is fed back to update the world prediction, so that both are recursively refined during inference. Rather than eliminating test-time world evolution altogether or rolling out the full future in pixel space, DAWN performs a short explicit latent rollout that is sufficient to support long-horizon trajectory generation in complex interactive scenes. Experiments show that DAWN achiev

Diffusion large language models (dLLMs) have shown advantages in text generation, particularly due to their inherent ability for parallel decoding. However, constrained by the quality--speed trade-off, existing inference solutions adopt conservative parallel strategies, leaving substantial efficiency potential underexplored. A core challenge is that parallel decoding assumes each position can be filled independently, but tokens are often semantically coupled. Thus, the correct choice at one position constrains valid choices at others. Without modeling these inter-token dependencies, parallel strategies produce deteriorated outputs. Motivated by this insight, we propose DAWN, a training-free, dependency-aware decoding method for fast dLLM inference. DAWN extracts token dependencies and leverages two key motivations: (1) positions dependent on unmasked certain positions become more reliable, (2) simultaneously unmasking strongly coupled uncertain positions induces errors. Given those findings, DAWN leverages a dependency graph to select more reliable unmasking positions at each iteration, achieving high parallelism with negligible loss in generation quality. Extensive experiments acr

Cosmic dawn 21-cm observations will extend standard-ruler cosmology into the first billion years, unlocking epochs inaccessible by the cosmic microwave background and large-scale structure. Realizing this promise requires an accurate model of the acoustic structure imprinted onto early star formation. At such early times, two counterbalancing phenomena -- matter overdensities and streaming velocities between cold dark matter and baryons -- modulate the spatial statistics of star formation. While overdensities dictate where early galaxy-bearing haloes form, regions of high velocity suppress star formation. Their combined influence on the intergalactic medium yields 21-cm fluctuations with both baryon (BAO) and velocity-induced (VAO) acoustic oscillations. Here we present the first prescription to decompose the 21-cm power spectrum into its constituent acoustic features. We find a percent-level offset between the BAO and VAO shapes which, if ignored in standard-ruler analyses, would bias inferred values of $H(z)$ by $\sim 2\%$; we provide a correction. Moreover, as the relative prominence of BAOs and VAOs ebbs and flows non-monotonically across cosmic dawn, we demonstrate how their e

The James Webb Space Telescope (JWST) has begun to revolutionize our view of the Cosmos. The discovery of Blue Monsters (i.e., ultra-compact yet very bright high-z galaxies) and the Little Red Dots (i.e., very compact dustless strong Balmer break cosmic dawn sources) pose significant challenges to pre-JWST era models of the assembly of first stars and galaxies. In addition, JWST data further strengthen the problem posed by the origin of the supermassive black holes that power the most distant quasars observed. Stars powered by Dark Matter annihilation (i.e., Dark Stars) can form out of primordial gas clouds during the cosmic dawn era and subsequently might grow via accretion and become supermassive. In this paper we argue that Supermassive Dark Stars (SMDSs) offer natural solutions to the three puzzles mentioned above.

In a recent paper we pointed out that supermassive black holes, with masses ranging from $10^5$ to $10^{10} M_\odot$ form naturally at cosmic dawn if the dark matter is QCD axions or axion-like particles with mass $m > 10^{-16}\, \mathrm{eV}/c^2$. No additional assumptions are required. Here we answer in detail the most commonly raised questions regarding our work.

The rapid evolution of Large Language Models (LLMs) has transformed them from basic conversational tools into sophisticated entities capable of complex reasoning and decision-making. These advancements have led to the development of specialized LLM-based agents designed for diverse tasks such as coding and web browsing. As these agents become more capable, the need for a robust framework that facilitates global communication and collaboration among them towards advanced objectives has become increasingly critical. Distributed Agents in a Worldwide Network (DAWN) addresses this need by offering a versatile framework that integrates LLM-based agents with traditional software systems, enabling the creation of agentic applications suited for a wide range of use cases. DAWN enables distributed agents worldwide to register and be easily discovered through Gateway Agents. Collaborations among these agents are coordinated by a Principal Agent equipped with reasoning strategies. DAWN offers three operational modes: No-LLM Mode for deterministic tasks, Copilot for augmented decision-making, and LLM Agent for autonomous operations. Additionally, DAWN ensures the safety and security of agent c

The enigmatic ultraviolet (UV) extinction bump at 2175 Angstrom, the strongest spectroscopic absorption feature superimposed on the interstellar extinction curve, has recently been detected at the cosmic dawn by the James Webb Space Telescope (JWST) in JADES-GS-z6-0, a distant galaxy at redshift z=6.71, corresponding to a cosmic age of just 800 million years after the Big Bang. Although small graphite grains have historically long been suggested as the carrier of the 2175 Angstrom extinction bump and graphite grains are expected to have already been pervasive in the early Universe, in this work we demonstrate that small graphite grains are not responsible for the UV extinction bump seen at the cosmic dawn in JADES-GS-z6-0, as the extinction bump arising from small graphite grains is too broad and peaks at wavelengths that are too short to be consistent with what is seen in JADES-GS-z6-0.

The James Webb Space Telescope (JWST) now allows us to observe galaxies at the end of cosmic dawn ($z \sim 10-15$) with unprecedented detail, revealing their morphologies, sizes, and internal structures. These observations offer crucial insights into the physical processes driving early galaxy formation. In this work, we introduce the Phoebos hydrodynamical cosmological simulation, a state-of-the-art 100 Mpc volume designed to study the formation and evolution of galaxies at the end of cosmic dawn and into the epoch of reionization. Phoebos includes a stellar feedback model that is intentionally weak, in order to address the high abundance of massive galaxies seen by JWST at early epochs. At variance with most large cosmological hydrodynamical simulations, we do not employ an effective equation of state model, instead our radiative cooling model allows us to capture the multi-phase nature of the gas inside and around galaxies. Phoebos reproduces key observables of early galaxy formation at $z \gtrsim 8$, including the stellar mass function and the stellar-to-halo mass relation. It also recovers the observed slope of the stellar size-to-mass relation and matches the specific star fo

Mapping out the first billion years using the 21-cm line with the Square Kilometer Array (SKA) will revolutionize our understanding of the cosmic dawn, reionization and the galaxies that drove these milestones. However, synergies with other telescopes in the form of cross correlations will be fundamental in making and confirming initial, low signal-to-noise claims of a detection. Participants of the 2023 European Southern Observatory (ESO) - SKA Observatory workshop discussed such synergies for Epoch of Reionization (EoR) and Cosmic Dawn (CD) science. Here we highlight some of the most promising candidates for cross-correlating SKA EoR/CD observations with ESO instruments such as the Multi-Object Optical and Near-infrared Spectrograph (MOONS), the MOSAIC multi-object spectrograph, and the ArmazoNes high Dispersion Echelle Spectrograph (ANDES).

Dark matter annihilation has the potential to leave an imprint on the properties of the first luminous structures at Cosmic Dawn as well as the overall evolution of the intergalactic medium (IGM). In this work, we employ a semi-analytic method to model dark matter annihilation during Cosmic Dawn (approximately redshift $z=20$ to $40$), examining potential modifications to IGM evolution as well as gas collapse, cooling, and star formation in mini-halos. Our analysis takes into account the effects of dark matter-baryon velocity offsets, utilizing the public 21cmvFAST code, and producing predictions for the 21cm global signal. The results from our simplified model suggest that dark matter annihilation can suppress the gas fraction in small halos and alter the molecular cooling process, while the impact on star formation might be positive or negative depending on parameters of the dark matter model as well as the redshift and assumptions about velocity offsets. This underscores the need for more comprehensive simulations of the effects of exotic energy injection at Cosmic Dawn as observational probes are providing us new insights into the process of reionization and the formation of fi

Talking head generation intends to produce vivid and realistic talking head videos from a single portrait and speech audio clip. Although significant progress has been made in diffusion-based talking head generation, almost all methods rely on autoregressive strategies, which suffer from limited context utilization beyond the current generation step, error accumulation, and slower generation speed. To address these challenges, we present DAWN (Dynamic frame Avatar With Non-autoregressive diffusion), a framework that enables all-at-once generation of dynamic-length video sequences. Specifically, it consists of two main components: (1) audio-driven holistic facial dynamics generation in the latent motion space, and (2) audio-driven head pose and blink generation. Extensive experiments demonstrate that our method generates authentic and vivid videos with precise lip motions, and natural pose/blink movements. Additionally, with a high generation speed, DAWN possesses strong extrapolation capabilities, ensuring the stable production of high-quality long videos. These results highlight the considerable promise and potential impact of DAWN in the field of talking head video generation. Fu

Motivated by the early excess of bright galaxies seen by JWST, we run zoom-in cosmological simulations of a massive galaxy at Cosmic Dawn, in a halo of $10^{11} M_\odot$ at $z = 9$, using the hydro-gravitational code RAMSES at an effective resolution $\sim 10~{\rm pc}$. We investigate physical mechanisms that enhance the star-formation efficiencies (SFEs) at the high gas densities of the star-forming regions in this galaxy ($\sim 3\times 10^3~{\rm cm^{-3}}$, $\sim 10^4~M_\odot/{\rm pc^2}$). Our fiducial star formation recipe uses a physically-motivated, turbulence-based, multi-freefall model, avoiding ad hoc extrapolation from lower redshifts. By $z = 9$, our simulated galaxy is a clumpy, thick, rotating disc with a high stellar mass $\sim 3\times 10^9~M_\odot$ and high star formation rate $\sim 50~M_\odot/{\rm yr}$. The high gas density makes supernova (SN) feedback less efficient, producing a high local SFE $\gtrsim 10\%$. The global SFE is set by feedback-driven outflows and only weakly correlated with the local SFE. Photoionization heating makes SN feedback more efficient, but the integrated SFE always remains high. Intense accretion at Cosmic Dawn seeds turbulence which reduce

The efficiency of radio emission is an important unknown parameter of early galaxies at cosmic dawn, as models with high efficiency have been shown to modify the cosmological 21-cm signal substantially, deepening the absorption trough and boosting the 21-cm power spectrum. Such models have been previously directly constrained by the overall extragalactic radio background as observed by ARCADE-2 and LWA-1. In this work, we constrain the clustering of high redshift radio sources by utilizing the observed upper limits on arcminute-scale anisotropy from the VLA at 4.9~GHz and ATCA at 8.7~GHz. Using a semi-numerical simulation of a plausible astrophysical model for illustration, we show that the clustering constraints on the radio efficiency are much stronger than those from the overall background intensity, by a factor that varies from 12 at redshift 7 to 30 at redshift 22. As a result, the predicted maximum depth of the global 21-cm signal is lowered by a factor of 5 (to 1700~mK), and the maximum 21-cm power spectrum peak at cosmic dawn is lowered by a factor of 24 (to $2\times 10^5$~mK$^2$). We conclude that the observed clustering is the strongest current direct constraint on such m

Recent discoveries of a significant population of bright galaxies at cosmic dawn $\left(z \gtrsim 10\right)$ have enabled critical tests of cosmological galaxy formation models. In particular, the bright end of the galaxy UV luminosity function (UVLF) appears higher than predicted by many models. Using approximately 25,000 galaxy snapshots at $8 \leq z \leq 12$ in a suite of FIRE-2 cosmological "zoom-in'' simulations from the Feedback in Realistic Environments (FIRE) project, we show that the observed abundance of UV-bright galaxies at cosmic dawn is reproduced in these simulations with a multi-channel implementation of standard stellar feedback processes, without any fine-tuning. Notably, we find no need to invoke previously suggested modifications such as a non-standard cosmology, a top-heavy stellar initial mass function, or a strongly enhanced star formation efficiency. We contrast the UVLFs predicted by bursty star formation in these original simulations to those derived from star formation histories (SFHs) smoothed over prescribed timescales (e.g., 100 Myr). The comparison demonstrates that the strongly time-variable SFHs predicted by the FIRE simulations play a key role in c

The shortest path problem is a typical problem in graph theory with wide potential applications. The state-of-the-art single-source shortest paths algorithm on the weight graph is the $Δ$-stepping algorithm, which can efficiently process weighted graphs in parallel. DAWN is an algorithm that addresses the shortest path problem on unweighted graphs, and we propose a weighted version that can handle graphs with weights edges, while maintaining the high scalability and parallelism features as DAWN. The novel version requires $O(μm)$ and $O(μ\cdot E_{wcc})$ times on the connected and unconnected graphs for SSSP problems, respectively. $E_{wcc}$ denote the number of edges included in the largest weakly connected component, and $μ$ is a constant denoting the average number of path transformations in the tasks. We tested the weighted version on the real graphs from Stanford Network Analysis Platform and SuiteSparse Matrix Collection, which outperformed the solution of $Δ$-stepping algorithm from Gunrock, achieving a speedup of 43.163$\times$.

Dawn is the first NASA mission to operate in the vicinity of the two most massive asteroids in the main belt, Ceres and Vesta. This double-rendezvous mission is enabled by the use of low-thrust solar electric propulsion. Dawn will arrive at Vesta in 2011 and will operate in its vicinity for approximately one year. Vesta's mass and non-spherical shape, coupled with its rotational period, presents very interesting challenges to a spacecraft that depends principally upon low-thrust propulsion for trajectory-changing maneuvers. The details of Vesta's high-order gravitational terms will not be determined until after Dawn's arrival at Vesta, but it is clear that their effect on Dawn operations creates the most complex operational environment for a NASA mission to date. Gravitational perturbations give rise to oscillations in Dawn's orbital radius, and it is found that trapping of the spacecraft is possible near the 1:1 resonance between Dawn's orbital period and Vesta's rotational period, located approximately between 520 and 580 km orbital radius.This resonant trapping can be escaped by thrusting at the appropriate orbital phase. Having passed through the 1:1 resonance, gravitational pe

The recent EDGES measurements of the global 21-cm signal from the cosmic dawn suggest that the kinetic temperature of the inter-galactic medium (IGM) might be significantly lower compared to its expected value. The colder IGM directly affects the hydrogen recombination of the universe during the cosmic dawn and dark ages by enhancing the rate of recombinations. Here, we study and quantify, the impact of the colder IGM scenario on the recombination history of the universe in the context of DM-baryonic interaction model which is widely used to explain the depth of the EDGES 21-cm signal. We find that, in general, the hydrogen ionisation fraction gets suppressed during the dark ages and cosmic dawn and the suppression gradually increases at lower redshifts until X-ray heating turns on. However, accurate estimation of the ionisation fraction requires knowledge of the entire thermal history of the IGM, from the epoch of thermal decoupling of hydrogen gas and the CMBR to the cosmic dawn. It is possible that two separate scenarios which predict very similar HI differential temperature during the cosmic dawn and are consistent with the EDGES 21-cm signal might have very different IGM tempe

The epoch of Cosmic Dawn, when the first stars and galaxies were born, is widely considered the final frontier of observational cosmology today. Mapping the period between Cosmic Dawn and the present-day provides access to more than 90% of the baryonic (normal) matter in the Universe, and unlocks several thousand times more Fourier modes of information than available in today's cosmological surveys. We review the progress in modelling baryonic gas observations as tracers of the cosmological large-scale structure from Cosmic Dawn to the present day. We illustrate how the description of dark matter haloes can be extended to describe baryonic gas abundances and clustering. This innovative approach allows us to fully utilize our current knowledge of astrophysics to constrain cosmological parameters from future observations. Combined with the information content of multi-messenger probes, this will also elucidate the properties of the first supermassive black holes at Cosmic Dawn. We present a host of fascinating implications for constraining physics beyond the $Λ$CDM model, including tests of the theories of inflation and the cosmological principle, the effects of non-standard dark mat

The nature of dark matter remains unknown, but upcoming measurements probing the high-redshift Universe may provide invaluable insight. In the presence of dark matter-baryon scattering, the suppression in the matter power spectrum and the colder mean gas temperature are expected to modify the evolution of cosmic dawn and reionization. However, the contributions from such interactions to the baryon and dark matter temperature perturbations have been neglected thus far. In this work, we derive these contributions, evolve the cosmological perturbations until the end of the dark ages and show that they may have a significant impact in the beginning of cosmic dawn. In particular, we find that the amplitude of the temperature power spectrum at large scales can change by up to 1--2 orders of magnitude and that the matter power spectrum is further suppressed with respect to $Λ$CDM by $5$-$10\%$ at $k\sim 200\, {\rm Mpc^{-1}}$ compared to the computation ignoring these contributions for scattering cross sections at current CMB limits. As a case example, we also compute the HI power spectrum from the dark ages, finding significant differences due to the changes in the temperature and ionizat