Search

We give a short and relatively elementary proof of the Hilton-Milner Theorem., Comment: 4 pages

Machine unlearning is a key requirement of many data protection regulations such as GDPR. Prior work on unlearning has mostly considered superficial unlearning tasks where a single or a few related pieces of information are required to be removed. However, the task of unlearning a fact is much more challenging in recent large language models (LLMs), because the facts in LLMs can be deduced from each other. In this work, we investigate whether current unlearning methods for LLMs succeed beyond superficial unlearning of facts. Specifically, we formally propose a framework and a definition for deep unlearning facts that are interrelated. We design the metric, recall, to quantify the extent of deep unlearning. To systematically evaluate deep unlearning, we construct a synthetic dataset EDU-RELAT, which consists of a synthetic knowledge base of family relationships and biographies, together with a realistic logical rule set that connects them. We use this dataset to test four unlearning methods in four LLMs at different sizes. Our findings reveal that in the task of deep unlearning only a single fact, they either fail to properly unlearn with high recall, or end up unlearning many other irrelevant facts. Our dataset and code are publicly available at: https://github.com/wrh14/deep_unlearning.

We propose two novel algorithms for constructing convex collision-free polytopes in robot configuration space. Finding these polytopes enables the application of stronger motion-planning frameworks such as trajectory optimization with Graphs of Convex Sets [1] and is currently a major roadblock in the adoption of these approaches. In this paper, we build upon IRIS-NP (Iterative Regional Inflation by Semidefinite & Nonlinear Programming) [2] to significantly improve tunability, runtimes, and scaling to complex environments. IRIS-NP uses nonlinear programming paired with uniform random initialization to find configurations on the boundary of the free configuration space. Our key insight is that finding near-by configuration-space obstacles using sampling is inexpensive and greatly accelerates region generation. We propose two algorithms using such samples to either employ nonlinear programming more efficiently (IRIS-NP2 ) or circumvent it altogether using a massively-parallel zero-order optimization strategy (IRIS-ZO). We also propose a termination condition that controls the probability of exceeding a user-specified permissible fraction-in-collision, eliminating a significant source of tuning difficulty in IRIS-NP. We compare performance across eight robot environments, showing that IRIS-ZO achieves an order-of-magnitude speed advantage over IRIS-NP. IRISNP2, also significantly faster than IRIS-NP, builds larger polytopes using fewer hyperplanes, enabling faster downstream computation. Website: https://sites.google.com/view/fastiris, Comment: 16 pages, 6 figures, accepted for publication in the proceedings of the International Symposium for Robotics Research 2024

The strong spin-orbit interaction in silicon and germanium hole quantum dots enables all-electric microwave control of single spins but is unsuited for multi-spin exchange-only qubits that rely on scalable discrete signals to suppress cross-talk and heating effects in large quantum processors. Here, we propose an exchange-only spin-orbit qubit that utilizes spin-orbit interactions to implement qubit gates and keeps the beneficial properties of the original encoding. Our encoding is robust to significant local variability in hole spin properties and, because it operates with two degenerate states, it eliminates the need for the rotating frame, avoiding the technologically demanding constraints of fast clocks and precise signal calibration. Unlike current exchange-only qubits, which require complex multi-step sequences prone to leakage, our qubit design enables low-leakage two-qubit gates in a single step, addressing critical challenges in scaling spin qubits.

Over the past several decades, time-series photometry of CSPNe has yielded significant results including, but not limited to, discoveries of nearly 100 binary systems, insights into pulsations and winds in young white dwarfs, and studies of stars undergoing very late thermal pulses. We have undertaken a systematic study of optical photometric variability of cataloged CSPNe, using the epochal photometric data from the Zwicky Transient Facility (ZTF). By applying appropriate variability metrics, we arrive at a list of 94 significantly variable CSPNe. Based on the timescales of the light-curve activity, we classify the variables broadly into short- and long-timescale variables. In this first paper in this series, we focus on the former, which is the majority class comprising 83 objects. We infer periods for six sources for the first time, and recover several known periodic variables. Among the aperiodic sources, most exhibit a jitter around a median flux with a stable amplitude, and a few show outbursts. We draw attention to WeSb 1, which shows a different kind of variability: prominent deep and aperiodic dips, resembling transits from a dust/debris disk. We find strong evidence for a binary nature of WeSb 1 (possibly an A- to G-type companion). The compactness of the emission lines and inferred high electron densities make WeSb 1 a candidate for either an EGB 6-type planetary nucleus, or a symbiotic system inside an evolved planetary nebula, both of which are rare objects. To demonstrate further promise with ZTF, we report three additional newly identified periodic sources that do not appear in the list of highly variable sources. Finally, we also introduce a two-dimensional metric space defined by the von Neumann statistics and Pearson Skew and demonstrate its effectiveness in identifying unique variables of astrophysical interest, like WeSb 1., Comment: 19 pages + 8 pages appendix, 5 tables, 17 figures; Submitted to PASP; Comments are welcome!

Hundreds of exoplanets between 1-1.8 times the size of the Earth have been discovered on close in orbits. However, these planets show such a diversity in densities that some appear to be made entirely of iron, while others appear to host gaseous envelopes. To test this diversity in composition, we update the masses of 5 rocky exoplanets (HD 93963 A b, Kepler-10 b, Kepler-100 b, Kepler-407 b, and TOI-1444 b) and present the confirmation of a new planet (TOI-1011) using 187 high precision RVs from Gemini/MAROON-X and Keck/KPF. Our updated planet masses suggest compositions closer to that of the Earth than previous literature values for all planets in our sample. In particular, we report that two previously identified ``super-Mercuries'' (Kepler-100 b and HD 93963 A b) have lower masses that suggest less iron-rich compositions. We then compare the ratio of iron to rock-building species to the abundance ratios of those elements in their host stars. These updated planet compositions do not suggest a steep relationship between planet and host star compositions, contradictory to previous results, and suggest that planets and host stars have similar abundance ratios., Comment: Submitted to AJ 09/30/2024

The Shortest-Path Problem in Graph of Convex Sets (SPP in GCS) is a recently developed optimization framework that blends discrete and continuous decision making. Many relevant problems in robotics, such as collision-free motion planning, can be cast and solved as an SPP in GCS, yielding lower-cost solutions and faster runtimes than state-of-the-art algorithms. In this paper, we are motivated by motion planning of robot arms that must operate swiftly in static environments. We consider a multi-query extension of the SPP in GCS, where the goal is to efficiently precompute optimal paths between given sets of initial and target conditions. Our solution consists of two stages. Offline, we use semidefinite programming to compute a coarse lower bound on the problem's cost-to-go function. Then, online, this lower bound is used to incrementally generate feasible paths by solving short-horizon convex programs. For a robot arm with seven joints, our method designs higher quality trajectories up to two orders of magnitude faster than existing motion planners., Comment: To appear in: The International Workshop on the Algorithmic Foundations of Robotics, WAFR 2024

Computational design of protein-binding proteins is a fundamental capability with broad utility in biomedical research and biotechnology. Recent methods have made strides against some target proteins, but on-demand creation of high-affinity binders without multiple rounds of experimental testing remains an unsolved challenge. This technical report introduces AlphaProteo, a family of machine learning models for protein design, and details its performance on the de novo binder design problem. With AlphaProteo, we achieve 3- to 300-fold better binding affinities and higher experimental success rates than the best existing methods on seven target proteins. Our results suggest that AlphaProteo can generate binders "ready-to-use" for many research applications using only one round of medium-throughput screening and no further optimization., Comment: 45 pages, 17 figures

We report the sky-projected spin-orbit angle $\lambda$ for HD 191939 b, the innermost planet in a 6 planet system, using Keck/KPF to detect the Rossiter-McLaughlin (RM) effect. Planet b is a sub-Neptune with radius 3.4 $\pm$ 0.8 R$_{\oplus}$ and mass 10.0 $\pm$ 0.7 M$_{\oplus}$ with an RM amplitude $<$1 ms$^{-1}$. We find the planet is consistent with a well-aligned orbit, measuring $\lambda= \, $ 3.7 $\pm$ 5.0 degrees. Additionally, we place new constraints on the mass and period of the distant super-Jupiter, planet f, finding it to be 2.88 $\pm$ 0.26 $M_J$ on a 2898 $\pm$ 152 day orbit. With these new orbital parameters, we perform a dynamical analysis of the system and constrain the mutual inclination of the non-transiting planet e to be smaller than 12 degrees relative to the plane shared by the inner three transiting planets. Additionally, the further planet f is inclined off this shared plane, the greater the amplitude of precession for the entire inner system, making it increasingly unlikely to measure an aligned orbit for planet b. Through this analysis, we show that this system's wide variety of planets are all well-aligned with the star and nearly co-planar, suggesting that the system formed dynamically cold and flat out of a well-aligned proto-planetary disk, similar to our own solar system., Comment: 11 pages, 4 figures, Accepted for publication in The Astronomical Journal

Attention is a key part of the transformer architecture. It is a sequence-to-sequence mapping that transforms each sequence element into a weighted sum of values. The weights are typically obtained as the softmax of dot products between keys and queries. Recent work has explored alternatives to softmax attention in transformers, such as ReLU and sigmoid activations. In this work, we revisit sigmoid attention and conduct an in-depth theoretical and empirical analysis. Theoretically, we prove that transformers with sigmoid attention are universal function approximators and benefit from improved regularity compared to softmax attention. Through detailed empirical analysis, we identify stabilization of large initial attention norms during the early stages of training as a crucial factor for the successful training of models with sigmoid attention, outperforming prior attempts. We also introduce FLASHSIGMOID, a hardware-aware and memory-efficient implementation of sigmoid attention yielding a 17% inference kernel speed-up over FLASHATTENTION2 on H100 GPUs. Experiments across language, vision, and speech show that properly normalized sigmoid attention matches the strong performance of softmax attention on a wide range of domains and scales, which previous attempts at sigmoid attention were unable to fully achieve. Our work unifies prior art and establishes best practices for sigmoid attention as a drop-in softmax replacement in transformers.

Efficient control schemes that enable fast, high-fidelity operations are essential for any practical quantum computation. However, current optimization protocols are intractable due to stringent requirements imposed by the microscopic systems encoding the qubit, including dense energy level spectra and cross talk, and generally require a trade-off between speed and fidelity of the operation. Here, we address these challenges by developing a general framework for optimal control based on the quantum metric tensor. This framework allows for fast and high-fidelity adiabatic pulses, even for a dense energy spectrum, based solely on the Hamiltonian of the system instead of the full time evolution propagator and independent of the size of the underlying Hilbert space. Furthermore, the framework suppresses diabatic transitions and state-dependent crosstalk effects without the need for additional control fields. As an example, we study the adiabatic charge transfer in a double quantum dot to find optimal control pulses with improved performance. We show that for the geometric protocol, the transfer fidelites are lower bounded $F>99\%$ for ultrafast 20 ns pulses, regardless of the size of the anti-crossing., Comment: 16 pages, 12 figures

The evolution of artificial intelligence (AI) presents an unprecedented opportunity for Historically Black Colleges and Universities (HBCUs) to lead in an era characterized by rapid technological advancement and societal transformation. This report explores the integral role HBCUs--with their history of academic excellence and commitment to inclusivity--can play in steering the AI revolution, ensuring that the Black community remains at the forefront of educational, economic, and social progress. Institutional leaders, faculty and instructors, and instructional support staff can benefit from the findings of this report, which is presented in the following sections: (1) Why AI Matters and the Unique Role of HBCUs in the AI Revolution; (2) An AI Policy and Practice Framework for Institutional Development; (3) AI and Curricular Innovation; (4) The Importance of Industry Partnerships and Student Development; and (5) AI at HBCUs: A Path to the Future. [This report was created in collaboration with the Thurgood Marshall College Fund.]

Additive manufacturing methods together with topology optimization have enabled the creation of multiscale structures with controlled spatially-varying material microstructure. However, topology optimization or inverse design of such structures in the presence of nonlinearities remains a challenge due to the expense of computational homogenization methods and the complexity of differentiably parameterizing the microstructural response. A solution to this challenge lies in machine learning techniques that offer efficient, differentiable mappings between the material response and its microstructural descriptors. This work presents a framework for designing multiscale heterogeneous structures with spatially varying microstructures by merging a homogenization-based topology optimization strategy with a consistent machine learning approach grounded in hyperelasticity theory. We leverage neural architectures that adhere to critical physical principles such as polyconvexity, objectivity, material symmetry, and thermodynamic consistency to supply the framework with a reliable constitutive model that is dependent on material microstructural descriptors. Our findings highlight the potential of integrating consistent machine learning models with density-based topology optimization for enhancing design optimization of heterogeneous hyperelastic structures under finite deformations.

Ensuring positioning integrity amid faulty measurements is crucial for safety-critical applications, making receiver autonomous integrity monitoring (RAIM) indispensable. This paper introduces a Bayesian RAIM algorithm with a streamlined architecture for snapshot-type 3D cellular positioning. Unlike traditional frequentist-type RAIM algorithms, it computes the exact posterior probability density function (PDF) of the position vector as a Gaussian mixture (GM) model using efficient message passing along a factor graph. This Bayesian approach retains all crucial information from the measurements, eliminates the need to discard faulty measurements, and results in tighter protection levels (PLs) in 3D space and 1D/2D subspaces that meet target integrity risk (TIR) requirements. Numerical simulations demonstrate that the Bayesian RAIM algorithm significantly outperforms a baseline algorithm, achieving over $50\%$ PL reduction at a comparable computational cost., Comment: Submitted to IEEE Transactions on Vehicular Technology

Hole spin qubits are rapidly emerging as the workhorse of semiconducting quantum processors because of their large spin-orbit interaction, enabling fast all-electric operations at low power. However, spin-orbit interaction also causes non-uniformities in devices, resulting in locally varying qubit energies and site-dependent anisotropies. While these anisotropies can be used to drive single-spins, if not properly harnessed, they can hinder the path toward large-scale quantum processors. Here, we report on a qubit in planar germanium, where we control the anisotropy of two spins in a double quantum dot. By characterising microwave-driven singlet-triplet qubits, we show two distinct operating regimes depending on the magnetic field direction. For in-plane fields, the two spins are largely anisotropic, enabling to measure all the available transitions; coherence times exceeding 3 $\mu$s are extracted. For out-of-plane fields, they have an isotropic response but preserve the substantial energy difference required to address the singlet-triplet qubit. Even in this field direction, where the qubit lifetime is strongly affected by nuclear spins, we find 400 ns coherence times. Our work adds a valuable tool to investigate and harness the anisotropy of spin qubits and can be implemented in any large-scale NxN device, facilitating the path towards scalable quantum processors.

Characterizing outer companions to hot Jupiters plays a crucial role in deciphering their origins. We present the discovery of a long-period giant planet, HD 118203 c ($m_{c}=11.79^{+0.69}_{-0.63}\ \mathrm{M_{J}}$, $a_{c}=6.28^{+0.10}_{-0.11}$ AU) exterior to a close-in eccentric hot Jupiter HD 118203 b ($P_{b}=6.135\ \mathrm{days}$, $m_{b}=2.14\pm{0.12}\ \mathrm{M_{J}}$, $r_{b}=1.14\pm{0.029}\ \mathrm{R_{J}}$, $e_{b}=0.31\pm{0.007}$) based on twenty-year radial velocities. Using Rossiter-McLaughlin (RM) observations from the Keck Planet Finder (KPF), we measured a low sky-projected spin-orbit angle $\lambda_{b}=-11^{\circ}.7^{+7.6}_{-10.0}$ for HD 118203 b and detected stellar oscillations in the host star, confirming its evolved status. Combining the RM observation with the stellar inclination measurement, we constrained the true spin-orbit angle of HD 118203 b as $\Psi_{b}<33^{\circ}.5\ (2\sigma)$, indicating the orbit normal of the hot Jupiter nearly aligned with the stellar spin axis. Furthermore, by combining radial velocities and Hipparcos-Gaia astrometric acceleration, we constrained the line-of-sight mutual inclination between the hot Jupiter and the outer planet to be $9^{\circ}.8^{+16.2}_{-9.3}$ at $2\sigma$ level. HD 118203 is one of first hot Jupiter systems where both the true spin-orbit angle of the hot Jupiter and the mutual inclination between inner and outer planets have been determined. Our results are consistent with a system-wide alignment, with low mutual inclinations between the outer giant planet, the inner hot Jupiter, and the host star. This alignment, along with the moderate eccentricity of HD 118203 c, implies that the system may have undergone coplanar high-eccentricity tidal migration. Under this framework, our dynamical analysis suggests an initial semi-major axis of 0.3 to 3.2 AU for the proto-hot Jupiter., Comment: 30 pages, 12 figures, accepted by AJ

Cataclysmic variables (CVs) are compact binary systems in which a white dwarf accretes matter from a Roche-lobe-filling companion star. In this study, we searched for new CVs in the Milky Way in the Chandra Source Catalog v2.0, cross-matched with Gaia Data Release 3 (DR3). We identified new CV candidates by combining X-ray and optical data in a color-color diagram called the ``X-ray Main Sequence". We used two different cuts in this diagram to compile pure and optically variable samples of CV candidates. We undertook optical spectroscopic follow-up observations with the Keck and Palomar Observatories to confirm the nature of these sources. We assembled a sample of 25,887 Galactic X-ray sources and found 14 new CV candidates. Seven objects show X-ray and/or optical variability. All sources show X-ray luminosity in the $\rm 10^{29}-10^{32}$ $\rm erg\ s^{-1}$ range, and their X-ray spectra can be approximated by a power-law model with photon indices in the $\rm \Gamma \sim 1-3$ range or an optically thin thermal emission model in the $\rm kT \sim 1-70$ keV range. We spectroscopically confirmed four CVs, discovering two new polars, one low accretion rate polar and a WZ~Sge-like low accretion rate CV. X-ray and optical properties of the other 9 objects suggest that they are also CVs (likely magnetic or dwarf novae), and one other object could be an eclipsing binary, but revealing their true nature requires further observations. These results show that a joint X-ray and optical analysis can be a powerful tool for finding new CVs in large X-ray and optical catalogs. X-ray observations such as those by Chandra are particularly efficient at discovering magnetic and low accretion rate CVs, which could be missed by purely optical surveys., Comment: 20 pages, 15 figures and 8 tables. Accepted for publication in Astronomy & Astrophysics

Modern artificial intelligence (AI) systems are powered by foundation models. This paper presents a new set of foundation models, called Llama 3. It is a herd of language models that natively support multilinguality, coding, reasoning, and tool usage. Our largest model is a dense Transformer with 405B parameters and a context window of up to 128K tokens. This paper presents an extensive empirical evaluation of Llama 3. We find that Llama 3 delivers comparable quality to leading language models such as GPT-4 on a plethora of tasks. We publicly release Llama 3, including pre-trained and post-trained versions of the 405B parameter language model and our Llama Guard 3 model for input and output safety. The paper also presents the results of experiments in which we integrate image, video, and speech capabilities into Llama 3 via a compositional approach. We observe this approach performs competitively with the state-of-the-art on image, video, and speech recognition tasks. The resulting models are not yet being broadly released as they are still under development.

The degree of alignment between a star's spin axis and the orbital plane of its planets (the stellar obliquity) is related to interesting and poorly understood processes that occur during planet formation and evolution. Hot Jupiters orbiting hot stars ($\gtrsim$6250 K) display a wide range of obliquities, while similar planets orbiting cool stars are preferentially aligned. Tidal dissipation is expected to be more rapid in stars with thick convective envelopes, potentially explaining this trend. Evolved stars provide an opportunity to test the damping hypothesis, particularly stars that were hot on the main sequence and have since cooled and developed deep convective envelopes. We present the first systematic study of the obliquities of hot Jupiters orbiting subgiants that recently developed convective envelopes using Rossiter-McLaughlin observations. Our sample includes two newly discovered systems in the Giants Transiting Giants Survey (TOI-6029 b, TOI-4379 b). We find that the orbits of hot Jupiters orbiting subgiants that have cooled below $\sim$6250 K are aligned or nearly aligned with the spin-axis of their host stars, indicating rapid tidal realignment after the emergence of a stellar convective envelope. We place an upper limit for the timescale of realignment for hot Jupiters orbiting subgiants at $\sim$500 Myr. Comparison with a simplified tidal evolution model shows that obliquity damping needs to be $\sim$4 orders of magnitude more efficient than orbital period decay to damp the obliquity without destroying the planet, which is consistent with recent predictions for tidal dissipation from inertial waves excited by hot Jupiters on misaligned orbits., Comment: 22 pages, 14 figures, 3 tables

We introduce the OATMEAL survey, an effort to measure the obliquities of stars with transiting brown dwarf companions. We observed a transit of the close-in ($P_{\rm orb} = 1.74 \,$ days) brown dwarf GPX-1 b using the Keck Planet Finder (KPF) spectrograph to measure the sky-projected angle between its orbital axis and the spin axis of its early F-type host star ($\lambda$). We measured $\lambda = 6.88 \pm 1.72 ^\circ$ (with additional unquantified systematic uncertainty), suggesting an orbit that is prograde and well aligned with the stellar equator. Hot Jupiters around early F stars are frequently found to have highly misaligned orbits, with polar and retrograde orbits being commonplace. It has been theorized that these misalignments stem from dynamical interactions, such as von Zeipel-Kozai-Lidov cycles, and are retained over long timescales due to weak tidal dissipation in stars with radiative envelopes. By comparing GPX-1 to similar systems under the frameworks of different tidal evolution theories, we argued that the rate of tidal dissipation is too slow to have re-aligned the system. This suggests that GPX-1 may have arrived at its close-in orbit via coplanar high-eccentricity migration or migration through an aligned protoplanetary disk. Our result for GPX-1 is one of few measurements of the obliquity of a star with a transiting brown dwarf. By enlarging the number of such measurements and comparing them with hot Jupiter systems, we will more clearly discern the differences between the mechanisms that dictate the formation and evolution of both classes of objects., Comment: 8 pages, 4 figures

Asteroseismology of dwarf stars cooler than the Sun is very challenging due to the low amplitudes and rapid timescales of oscillations. Here, we present the asteroseismic detection of solar-like oscillations at 4-minute timescales ($\nu_{\mathrm{max}}\sim4300\mu$Hz) in the nearby K-dwarf $\sigma$ Draconis using extreme precision Doppler velocity observations from the Keck Planet Finder and 20-second cadence photometry from NASA's Transiting Exoplanet Survey Satellite. The star is the coolest dwarf star to date with both velocity and luminosity observations of solar-like oscillations, having amplitudes of $5.9\pm0.8\,$cm$\,\text{s}^{-1}$ and $0.8\pm0.2$ ppm, respectively. These measured values are in excellent agreement with established luminosity-velocity amplitude relations for oscillations and provide further evidence that mode amplitudes for stars with $T_{\mathrm{eff}}<\,5500\,$K diminish in scale following a $(L/M)^{1.5}$ relation. By modeling the star's oscillation frequencies from photometric data, we measure an asteroseismic age of $4.5\pm0.9\,\rm{(ran)} \pm 1.2\,\rm{(sys)}$ Gyr. The observations demonstrate the capability of next-generation spectrographs and precise space-based photometry to extend observational asteroseismology to nearby cool dwarfs, which are benchmarks for stellar astrophysics and prime targets for directly imaging planets using future space-based telescopes., Comment: Accepted for publication in The Astrophysical Journal

We present the first spectroscopic transit results from the newly commissioned Keck Planet Finder on the Keck-I telescope at W. M. Keck Observatory. We observed a transit of KELT-18 b, an inflated ultra-hot Jupiter orbiting a hot star ($T_\text{eff} = 6670$ K) with a binary stellar companion. By modeling the perturbation to the measured cross correlation functions using the Reloaded Rossiter-McLaughlin technique, we derived a sky projected obliquity of $\lambda = -94.8 \pm 0.7$ deg ($\psi = 93.8_{-1.8}^{+1.6}$ deg for isotropic $i_\star$). The data are consistent with an extreme stellar differential rotation ($\alpha = 0.9$), though a more likely explanation is moderate center-to-limb variations of the emergent stellar spectrum. We see additional evidence for the latter from line widths increasing towards the limb. Using loose constraints on the stellar rotation period from observed variability in the available TESS photometry, we were able to constrain the stellar inclination and thus the true 3D stellar obliquity to $\psi = 91.7_{-1.8}^{+2.2}$ deg. KELT-18 b could have obtained its polar orbit through high-eccentricity migration initiated by Kozai-Lidov oscillations induced by the binary stellar companion KELT-18 B, as the two likely have a large mutual inclination as evidenced by Gaia astrometry. KELT-18 b adds another data point to the growing population of close-in polar planets, particularly around hot stars., Comment: 17 pages, 8 figures, submitted to AJ (in revision)

The orbits of close-in exoplanets provide clues to their formation and evolutionary history. Many close-in exoplanets likely formed far out in their protoplanetary disks and migrated to their current orbits, perhaps via high-eccentricity migration (HEM), a process that can also excite obliquities. A handful of known exoplanets are perhaps caught in the act of HEM, as they are observed on highly eccentric orbits with tidal circularization timescales shorter than their ages. One such exoplanet is Kepler-1656 b, which is also the only known non-giant exoplanet (<100 $M_\oplus$) with an extreme eccentricity (e=0.84). We measured the sky-projected obliquity of Kepler-1656 b by observing the Rossiter-McLaughlin effect during a transit with the Keck Planet Finder. Our data are consistent with an aligned orbit, but are also consistent with moderate misalignment with $\lambda < 50$ deg at 95% confidence, with the most likely solution of $35^{+14.9}_{-21.6}$ deg. A low obliquity would be an unlikely outcome of most eccentricity-exciting scenarios, but we show that the properties of the outer companion in the system are consistent with the coplanar HEM mechanism. Alternatively, if the system is not relatively coplanar (<20 deg mutual inclination), Kepler-1656 b may be presently at a rare snapshot of long-lived eccentricity oscillations that do not induce migration. Kepler-1656 b is only the fourth exoplanet with e>0.8 to have its obliquity constrained; expanding this population will help establish the degree to which orbital misalignment accompanies migration. Future work that constrains the mutual inclinations of outer perturbers will be key for distinguishing plausible mechanisms., Comment: 11 pages, 5 figures, accepted to ApJL

TOI-6255~b (GJ 4256) is an Earth-sized planet (1.079$\pm0.065$ $R_\oplus$) with an orbital period of only 5.7 hours. With the newly commissioned Keck Planet Finder (KPF) and CARMENES spectrographs, we determined the planet's mass to be 1.44$\pm$0.14 $M_{\oplus}$. The planet is just outside the Roche limit, with $P_{\rm orb}/P_{\rm Roche}$ = 1.13 $\pm0.10$. The strong tidal force likely deforms the planet into a triaxial ellipsoid with a long axis that is $\sim$10\% longer than the short axis. Assuming a reduced stellar tidal quality factor $Q_\star^\prime \approx10^7$, we predict that tidal orbital decay will cause TOI-6255 to reach the Roche limit in roughly 400 Myr. Such tidal disruptions may produce the possible signatures of planet engulfment that have been on stars with anomalously high refractory elemental abundances compared to its conatal binary companion. TOI-6255 b is also a favorable target for searching for star-planet magnetic interactions, which might cause interior melting and hasten orbital decay. TOI-6255 b is a top target (Emission Spectroscopy Metric of about 24) for phase curve observations with the James Webb Space Telescope., Comment: 18 pages, 7 figures, 5 tables, accepted to AAS Journals. The first RV mass measurement from the Keck Planet Finder

The Internet of Things (IoT) network integrating billions of smart physical devices embedded with sensors, software, and communication technologies is a critical and rapidly expanding component of our modern world. The IoT ecosystem provides a rich source of real-world modalities such as motion, thermal, geolocation, imaging, depth, sensors, and audio to recognize the states of humans and physical objects. Machine learning presents a rich opportunity to automatically process IoT data at scale, enabling efficient inference for understanding human wellbeing, controlling physical devices, and interconnecting smart cities. To realize this potential, we introduce IoT-LM, an open-source large multisensory language model tailored for the IoT ecosystem. IoT-LM is enabled by two technical contributions: the first is MultiIoT, the most expansive unified IoT dataset to date, encompassing over 1.15 million samples from 12 modalities and 8 tasks prepared for multisensory pre-training and instruction-tuning. The second is a new multisensory multitask adapter layer to condition pre-trained large language models on multisensory IoT data. Not only does IoT-LM yield substantial improvements on 8 supervised IoT classification tasks, but it also demonstrates new interactive question-answering, reasoning, and dialog capabilities conditioned on IoT sensors. We release IoT-LM's data sources and new multisensory language modeling framework., Comment: arXiv admin note: text overlap with arXiv:2311.06217

We consider large-scale, implicit-search-based solutions to Shortest Path Problems on Graphs of Convex Sets (GCS). We propose GCS*, a forward heuristic search algorithm that generalizes A* search to the GCS setting, where a continuous-valued decision is made at each graph vertex, and constraints across graph edges couple these decisions, influencing costs and feasibility. Such mixed discrete-continuous planning is needed in many domains, including motion planning around obstacles and planning through contact. This setting provides a unique challenge for best-first search algorithms: the cost and feasibility of a path depend on continuous-valued points chosen along the entire path. We show that by pruning paths that are cost-dominated over their entire terminal vertex, GCS* can search efficiently while still guaranteeing cost-optimality and completeness. To find satisficing solutions quickly, we also present a complete but suboptimal variation, pruning instead reachability-dominated paths. We implement these checks using polyhedral-containment or sampling-based methods. The former implementation is complete and cost-optimal, while the latter is probabilistically complete and asymptotically cost-optimal and performs effectively even with minimal samples in practice. We demonstrate GCS* on planar pushing tasks where the combinatorial explosion of contact modes renders prior methods intractable and show it performs favorably compared to the state-of-the-art. Project website: https://shaoyuan.cc/research/gcs-star/, Comment: Accepted to WAFR 2024. Conference Ready Version

This paper presents Diffusion Forcing, a new training paradigm where a diffusion model is trained to denoise a set of tokens with independent per-token noise levels. We apply Diffusion Forcing to sequence generative modeling by training a causal next-token prediction model to generate one or several future tokens without fully diffusing past ones. Our approach is shown to combine the strengths of next-token prediction models, such as variable-length generation, with the strengths of full-sequence diffusion models, such as the ability to guide sampling to desirable trajectories. Our method offers a range of additional capabilities, such as (1) rolling-out sequences of continuous tokens, such as video, with lengths past the training horizon, where baselines diverge and (2) new sampling and guiding schemes that uniquely profit from Diffusion Forcing's variable-horizon and causal architecture, and which lead to marked performance gains in decision-making and planning tasks. In addition to its empirical success, our method is proven to optimize a variational lower bound on the likelihoods of all subsequences of tokens drawn from the true joint distribution. Project website: https://boyuan.space/diffusion-forcing, Comment: Project website: https://boyuan.space/diffusion-forcing Code: https://github.com/buoyancy99/diffusion-forcing

Recent work of Scott and Wilmer and of Woodroofe extends the Erd\H{o}s-Ko-Rado theorem from set systems to subspaces of k-forms in an exterior algebra. We prove an extension of the Hilton-Milner theorem to the exterior algebra setting, answering in a strong way a question asked by these authors., Comment: 15 pages

We present the first high-resolution, high-frequency radio continuum survey that fully maps an extragalactic deep field: the 10GHz survey of the Great Observatories Origins Deep Survey-North (GOODS-N) field. This is a Large Program of the Karl G. Jansky Very Large Array that allocated 380 hours of observations using the X-band ($8-12$GHz) receivers, leading to a 10GHz mosaic of the GOODS-field with an average rms noise $\sigma_{\rm n}=671\,\rm nJy\,beam^{-1}$ and angular resolution $\theta_{1/2}=0.22$arcsec across 297$\rm arcmin^2$. To maximize the brightness sensitivity we also produce a low-resolution mosaic with $\theta_{1/2}=1.0$arcsec and $\sigma_{\rm n}=968\,\rm nJy\,beam^{-1}$, from which we derive our master catalog containing 256 radio sources detected with peak signal-to-noise ratio $\geq 5$. Radio source size and flux density estimates from the high-resolution mosaic are provided in the master catalog as well. The total fraction of spurious sources in the catalog is 0.75%. Monte Carlo simulations are performed to derive completeness corrections of the catalog. We find that the 10GHz radio source counts in the GOODS-N field agree, in general, with predictions from numerical simulations/models and expectations from 1.4 and 3GHz radio counts., Comment: Accepted for publication in ApJ

There is an ongoing debate on balancing the benefits and risks of artificial intelligence (AI) as AI is becoming critical to improving healthcare delivery and patient outcomes. Such improvements are essential in resource-constrained settings where millions lack access to adequate healthcare services, such as in Africa. AI in such a context can potentially improve the effectiveness, efficiency, and accessibility of healthcare services. Nevertheless, the development and use of AI-driven healthcare systems raise numerous ethical, legal, and socio-economic issues. Justice is a major concern in AI that has implications for amplifying social inequities. This paper discusses these implications and related justice concepts such as solidarity, Common Good, sustainability, AI bias, and fairness. For Africa to effectively benefit from AI, these principles should align with the local context while balancing the risks. Compared to mainstream ethical debates on justice, this perspective offers context-specific considerations for equitable healthcare AI development in Africa., Comment: 14 pages

Data are essential in developing healthcare artificial intelligence (AI) systems. However, patient data collection, access, and use raise ethical concerns, including informed consent, data bias, data protection and privacy, data ownership, and benefit sharing. Various ethical frameworks have been proposed to ensure the ethical use of healthcare data and AI, however, these frameworks often align with Western cultural values, social norms, and institutional contexts emphasizing individual autonomy and well-being. Ethical guidelines must reflect political and cultural settings to account for cultural diversity, inclusivity, and historical factors such as colonialism. Thus, this paper discusses healthcare data ethics in the AI era in Africa from the Ubuntu philosophy perspective. It focuses on the contrast between individualistic and communitarian approaches to data ethics. The proposed framework could inform stakeholders, including AI developers, healthcare providers, the public, and policy-makers about healthcare data ethical usage in AI in Africa., Comment: 16 pages

Large policies pretrained on a combination of Internet-scale vision-language data and diverse robot demonstrations have the potential to change how we teach robots new skills: rather than training new behaviors from scratch, we can fine-tune such vision-language-action (VLA) models to obtain robust, generalizable policies for visuomotor control. Yet, widespread adoption of VLAs for robotics has been challenging as 1) existing VLAs are largely closed and inaccessible to the public, and 2) prior work fails to explore methods for efficiently fine-tuning VLAs for new tasks, a key component for adoption. Addressing these challenges, we introduce OpenVLA, a 7B-parameter open-source VLA trained on a diverse collection of 970k real-world robot demonstrations. OpenVLA builds on a Llama 2 language model combined with a visual encoder that fuses pretrained features from DINOv2 and SigLIP. As a product of the added data diversity and new model components, OpenVLA demonstrates strong results for generalist manipulation, outperforming closed models such as RT-2-X (55B) by 16.5% in absolute task success rate across 29 tasks and multiple robot embodiments, with 7x fewer parameters. We further show that we can effectively fine-tune OpenVLA for new settings, with especially strong generalization results in multi-task environments involving multiple objects and strong language grounding abilities, and outperform expressive from-scratch imitation learning methods such as Diffusion Policy by 20.4%. We also explore compute efficiency; as a separate contribution, we show that OpenVLA can be fine-tuned on consumer GPUs via modern low-rank adaptation methods and served efficiently via quantization without a hit to downstream success rate. Finally, we release model checkpoints, fine-tuning notebooks, and our PyTorch codebase with built-in support for training VLAs at scale on Open X-Embodiment datasets., Comment: Website: https://openvla.github.io/

The computational power and fault-tolerance of future large-scale quantum processors derive in large part from the connectivity between the qubits. One approach to increase connectivity is to engineer qubit-qubit interactions at a distance. Alternatively, the connectivity can be increased by physically displacing the qubits. This has been explored in trapped-ion experiments and using neutral atoms trapped with optical tweezers. For semiconductor spin qubits, several studies have investigated spin coherent shuttling of individual electrons, but high-fidelity transport over extended distances remains to be demonstrated. Here we report shuttling of an electron inside an isotopically purified Si/SiGe heterostructure using electric gate potentials. First, we form static quantum dots, and study how spin coherence decays as we repeatedly move a single electron between up to five dots. Next, we create a traveling wave potential to transport an electron in a moving quantum dot. This second method shows substantially better spin coherence than the first. It allows us to displace an electron over an effective distance of 10 {\mu}m in under 200 ns with an average fidelity of 99%. These results will guide future efforts to realize large-scale semiconductor quantum processors, making use of electron shuttling both within and between qubit arrays., Comment: 15 pages, 15 figures

Commercially impactful quantum algorithms such as quantum chemistry and Shor's algorithm require a number of qubits and gates far beyond the capacity of any existing quantum processor. Distributed architectures, which scale horizontally by networking modules, provide a route to commercial utility and will eventually surpass the capability of any single quantum computing module. Such processors consume remote entanglement distributed between modules to realize distributed quantum logic. Networked quantum computers will therefore require the capability to rapidly distribute high fidelity entanglement between modules. Here we present preliminary demonstrations of some key distributed quantum computing protocols on silicon T centres in isotopically-enriched silicon. We demonstrate the distribution of entanglement between modules and consume it to apply a teleported gate sequence, establishing a proof-of-concept for T centres as a distributed quantum computing and networking platform., Comment: 14 pages, 13 figures

We prove a Faber-Krahn inequality for the Laplacian with drift under Robin boundary condition, provided that the $\beta$ parameter in the Robin condition is large enough. The proof relies on a compactness argument, on the convergence of Robin eigenvalues to Dirichlet eigenvalues when $\beta$ goes to infinity, and on a strict Faber-Krahn inequality under Dirichlet boundary condition. We also show the existence and uniqueness of drifts $v$ satisfying some $L^\infty$ constraints and minimizing or maximizing the principal eigenvalue of $-\Delta+v\cdot\nabla$ in a fixed domain and with a fixed parameter $\beta>0$ in the Robin condition., Comment: 16 pages

During the first half of the fourth observing run (O4a) of the International Gravitational Wave Network (IGWN), the Zwicky Transient Facility (ZTF) conducted a systematic search for kilonova (KN) counterparts to binary neutron star (BNS) and neutron star-black hole (NSBH) merger candidates. Here, we present a comprehensive study of the five high-significance (FAR < 1 per year) BNS and NSBH candidates in O4a. Our follow-up campaigns relied on both target-of-opportunity observations (ToO) and re-weighting of the nominal survey schedule to maximize coverage. We describe the toolkit we have been developing, Fritz, an instance of SkyPortal, instrumental in coordinating and managing our telescope scheduling, candidate vetting, and follow-up observations through a user-friendly interface. ZTF covered a total of 2841 deg$^2$ within the skymaps of the high-significance GW events, reaching a median depth of g~20.2 mag. We circulated 15 candidates, but found no viable KN counterpart to any of the GW events. Based on the ZTF non-detections of the high-significance events in O4a, we used a Bayesian approach, nimbus, to quantify the posterior probability of KN model parameters that are consistent with our non-detections. Our analysis favors KNe with initial absolute magnitude fainter than -16 mag. The joint posterior probability of a GW170817-like KN associated with all our O4a follow-ups was 64%. Additionally, we use a survey simulation software, simsurvey, to determine that our combined filtered efficiency to detect a GW170817-like KN is 36%, when considering the 5 confirmed astrophysical events in O3 (1 BNS and 4 NSBH), along with our O4a follow-ups. Following Kasliwal et al. (2020), we derived joint constraints on the underlying KN luminosity function based on our O3 and O4a follow-ups, determining that no more than 76% of KNe fading at 1 mag/day can peak at a magnitude brighter than -17.5 mag., Comment: submitted

We present the tidal disruption event (TDE) AT2022lri, hosted in a nearby ($\approx\!144$ Mpc) quiescent galaxy with a low-mass massive black hole ($10^4\,M_\odot < M_{\rm BH} < 10^6\,M_\odot$). AT2022lri belongs to the TDE-H+He subtype. More than 1 Ms of X-ray data were collected with NICER, Swift, and XMM-Newton from 187 d to 672 d after peak. The X-ray luminosity gradually declined from $1.5\times 10^{44}\,{\rm erg\,s^{-1}}$ to $1.5\times 10^{43}\,{\rm erg\,s^{-1}}$ and remains much above the UV and optical luminosity, consistent with a super-Eddington accretion flow viewed face-on. Sporadic strong X-ray dips atop a long-term decline are observed, with variability timescale of $\approx\!0.5$ hr--1 d and amplitude of $\approx\!2$--8. When fitted with simple continuum models, the X-ray spectrum is dominated by a thermal disk component with inner temperature going from $\sim\! 146$ eV to $\sim\! 86$ eV. However, there are residual features that peak around 1 keV, which, in some cases, cannot be reproduced by a single broad emission line. We analyzed a subset of time-resolved spectra with two physically motivated models describing either a scenario where ionized absorbers contribute extra absorption and emission lines or where disk reflection plays an important role. Both models provide good and statistically comparable fits, show that the X-ray dips are correlated with drops in the inner disk temperature, and require the existence of sub-relativistic (0.1--0.3$c$) ionized outflows. We propose that the disk temperature fluctuation stems from episodic drops of the mass accretion rate triggered by magnetic instabilities or/and wobbling of the inner accretion disk along the black hole's spin axis., Comment: 35 pages, 20 figures, submitted

Inherent bias within society can be amplified and perpetuated by artificial intelligence (AI) systems. To address this issue, a wide range of solutions have been proposed to identify and mitigate bias and enforce fairness for individuals and groups. Recently, Graph Laplacian Regularizer (GLR), a regularization technique from the semi-supervised learning literature has been used as a substitute for the common Lipschitz condition to enhance individual fairness. Notable prior work has shown that enforcing individual fairness through a GLR can improve the transfer learning accuracy of AI models under covariate shifts. However, the prior work defines a GLR on the source and target data combined, implicitly assuming that the target data are available at train time, which might not hold in practice. In this work, we investigated whether defining a GLR independently on the train and target data could maintain similar accuracy. Furthermore, we introduced the Normalized Fairness Gain score (NFG) to measure individual fairness by measuring the amount of gained fairness when a GLR is used versus not. We evaluated the new and original methods under NFG, the Prediction Consistency (PC), and traditional classification metrics on the German Credit Approval dataset. The results showed that the two models achieved similar statistical mean performances over five-fold cross-validation. Furthermore, the proposed metric showed that PC scores can be misleading as the scores can be high and statistically similar to fairness-enhanced models while NFG scores are small. This work therefore provides new insights into when a GLR effectively enhances individual fairness and the pitfalls of PC., Comment: 14 pages, 1 figure, and 2 tables

The introduction of genome engineering technology has transformed biomedical research, making it possible to make precise changes to genetic information. However, creating an efficient gene-editing system requires a deep understanding of CRISPR technology, and the complex experimental systems under investigation. While Large Language Models (LLMs) have shown promise in various tasks, they often lack specific knowledge and struggle to accurately solve biological design problems. In this work, we introduce CRISPR-GPT, an LLM agent augmented with domain knowledge and external tools to automate and enhance the design process of CRISPR-based gene-editing experiments. CRISPR-GPT leverages the reasoning ability of LLMs to facilitate the process of selecting CRISPR systems, designing guide RNAs, recommending cellular delivery methods, drafting protocols, and designing validation experiments to confirm editing outcomes. We showcase the potential of CRISPR-GPT for assisting non-expert researchers with gene-editing experiments from scratch and validate the agent's effectiveness in a real-world use case. Furthermore, we explore the ethical and regulatory considerations associated with automated gene-editing design, highlighting the need for responsible and transparent use of these tools. Our work aims to bridge the gap between beginner biological researchers and CRISPR genome engineering techniques, and demonstrate the potential of LLM agents in facilitating complex biological discovery tasks.

Learning-based neural network (NN) control policies have shown impressive empirical performance in a wide range of tasks in robotics and control. However, formal (Lyapunov) stability guarantees over the region-of-attraction (ROA) for NN controllers with nonlinear dynamical systems are challenging to obtain, and most existing approaches rely on expensive solvers such as sums-of-squares (SOS), mixed-integer programming (MIP), or satisfiability modulo theories (SMT). In this paper, we demonstrate a new framework for learning NN controllers together with Lyapunov certificates using fast empirical falsification and strategic regularizations. We propose a novel formulation that defines a larger verifiable region-of-attraction (ROA) than shown in the literature, and refines the conventional restrictive constraints on Lyapunov derivatives to focus only on certifiable ROAs. The Lyapunov condition is rigorously verified post-hoc using branch-and-bound with scalable linear bound propagation-based NN verification techniques. The approach is efficient and flexible, and the full training and verification procedure is accelerated on GPUs without relying on expensive solvers for SOS, MIP, nor SMT. The flexibility and efficiency of our framework allow us to demonstrate Lyapunov-stable output feedback control with synthesized NN-based controllers and NN-based observers with formal stability guarantees, for the first time in literature. Source code at https://github.com/Verified-Intelligence/Lyapunov_Stable_NN_Controllers, Comment: Paper accepted by ICML 2024

The process of identifying a compound from its mass spectrum is a critical step in the analysis of complex mixtures. Typical solutions for the mass spectrum to compound (MS2C) problem involve matching the unknown spectrum against a library of known spectrum-molecule pairs, an approach that is limited by incomplete library coverage. Compound to mass spectrum (C2MS) models can improve retrieval rates by augmenting real libraries with predicted spectra. Unfortunately, many existing C2MS models suffer from problems with prediction resolution, scalability, or interpretability. We develop a new probabilistic method for C2MS prediction, FraGNNet, that can efficiently and accurately predict high-resolution spectra. FraGNNet uses a structured latent space to provide insight into the underlying processes that define the spectrum. Our model achieves state-of-the-art performance in terms of prediction error, and surpasses existing C2MS models as a tool for retrieval-based MS2C., Comment: 21 pages, 4 figures, 9 tables

Suppose $\Delta$ is a pure simplicial complex on $n$ vertices having dimension $d$ and let $c = n-d-1$ be its codimension in the simplex. Terai and Yoshida proved that if the number of facets of $\Delta$ is at least $\binom{n}{c}-2c+1$, then $\Delta$ is Cohen-Macaulay. We improve this result by showing that these hypotheses imply the stronger condition that $\Delta$ is vertex decomposable. We give examples to show that this bound is optimal, and that the conclusion cannot be strengthened to the class of matroids or shifted complexes. We explore an application to Simon's Conjecture and discuss connections to other results from the literature., Comment: In this version, included a connection to geometric vertex decomposability in the end. Modified the statement of Lemma 3.3, and proof of Theorem 1, along with other minor changes. Set to appear in the Electronic Journal of Combinatorics

Contrastive learning typically matches pairs of related views among a number of unrelated negative views. Views can be generated (e.g. by augmentations) or be observed. We investigate matching when there are more than two related views which we call poly-view tasks, and derive new representation learning objectives using information maximization and sufficient statistics. We show that with unlimited computation, one should maximize the number of related views, and with a fixed compute budget, it is beneficial to decrease the number of unique samples whilst increasing the number of views of those samples. In particular, poly-view contrastive models trained for 128 epochs with batch size 256 outperform SimCLR trained for 1024 epochs at batch size 4096 on ImageNet1k, challenging the belief that contrastive models require large batch sizes and many training epochs., Comment: Accepted to ICLR 2024. 42 pages, 7 figures, 3 tables, loss pseudo-code included in appendix

This work is devoted to the study of so-called ``reverse Riesz'' inequalities and suitable variants in the context of some fractal-like cable systems. It was already proved by L. Chen, T. Coulhon, J. Feneuil and the second author that, in the Vicsek cable system, the inequality $\left\Vert \Delta^{1/2}f\right\Vert_p\lesssim \left\Vert \nabla f\right\Vert_p$ is false for all $p\in [1,2)$. Following a recent joint paper by the two authors and M. Yang, we examine the validity of ``reverse quasi-Riesz'' inequalities, of the form $\left\Vert \Delta^{\gamma}e^{-\Delta}f\right\Vert_p\lesssim \left\Vert \nabla f\right\Vert_p$, in the (unbounded) Vicsek cable system, for $p\in (1,+\infty)$ and $\gamma>0$. These reverse inequalities are strongly related to the problem of $L^p$ boundedness of the operators $\nabla e^{-\Delta}\Delta^{-\varepsilon}$, the so-called ``quasi-Riesz transforms'' (at infinity), introduced by L. Chen in her PhD thesis. Our main result is an almost complete characterization of the sets of $\gamma\in (0,1)$ and $p\in (1,+\infty)$ such that the reverse quasi-Riesz inequality holds in the Vicsek cable system. It remains an open question to investigate reverse quasi-Riesz inequalities for other cable systems, or for manifolds built out of these.