arXiv subjects

Christopher Anderson

Publications and source records attributed to Christopher Anderson.

4 recordsLinked to original sources

arXiv2026

Fauna Sprout: A lightweight, approachable, developer-ready humanoid robot

Recent advances in learned control, large-scale simulation, and generative models have accelerated progress toward general-purpose robotic controllers, yet the field still lacks platforms suitable for safe, expressive, long-term deployment in human environments. Most existing humanoids are either closed industrial systems or academic prototypes that are difficult to deploy and operate around people, limiting progress in robotics. We introduce Sprout, a developer platform designed to address these limitations through an emphasis on safety, expressivity, and developer accessibility. Sprout adopts a lightweight form factor with compliant control, limited joint torques, and soft exteriors to support safe operation in shared human spaces. The platform integrates whole-body control, manipulation with integrated grippers, and virtual-reality-based teleoperation within a unified hardware-software stack. An expressive head further enables social interaction -- a domain that remains underexplored on most utilitarian humanoids. By lowering physical and technical barriers to deployment, Sprout expands access to capable humanoid platforms and provides a practical basis for developing embodied intelligence in real human environments.

cs.RO↗

arXiv2025

Generalization is not a universal guarantee: Estimating similarity to training data with an ensemble out-of-distribution metric

Failure of machine learning models to generalize to new data is a core problem limiting the reliability of AI systems, partly due to the lack of simple and robust methods for comparing new data to the original training dataset. We propose a standardized approach for assessing data similarity in a model-agnostic manner by constructing a supervised autoencoder for generalizability estimation (SAGE). We compare points in a low-dimensional embedded latent space, defining empirical probability measures for k-Nearest Neighbors (kNN) distance, reconstruction of inputs and task-based performance. As proof of concept for classification tasks, we use MNIST and CIFAR-10 to demonstrate how an ensemble output probability score can separate deformed images from a mixture of typical test examples, and how this SAGE score is robust to transformations of increasing severity. As further proof of concept, we extend this approach to a regression task using non-imaging data (UCI Abalone). In all cases, we show that out-of-the-box model performance increases after SAGE score filtering, even when applied to data from the model's own training and test datasets. Our out-of-distribution scoring method can be introduced during several steps of model construction and assessment, leading to future improvements in responsible deep learning implementation.

cs.LG↗

arXiv2024

The Radio and Microwave Sky as Seen by Juno on its Mission to Jupiter

We present six nearly full-sky maps made from data taken by radiometers on the Juno satellite during its 5-year flight to Jupiter. The maps represent integrated emission over $\sim 4\%$ passbands spaced approximately in octaves between 600 MHz and 21.9 GHz. Long time-scale offset drifts are removed in all bands, and, for the two lowest frequency bands, gain drifts are also removed from the maps via a self-calibration algorithm similar to the NPIPE pipeline used by the Planck collaboration. We show that, after this solution is applied, statistical noise in the maps is consistent with thermal radiometer noise and expected levels of correlated noise on the gain and noise drift solutions. We verify our map solutions with several consistency tests and end-to-end simulations. We also estimate the level of systematic pixelization noise and polarization leakage via simulations.

astro-ph.IM↗

arXiv2021

Overview and status of EXCLAIM, the experiment for cryogenic large-aperture intensity mapping

The EXperiment for Cryogenic Large-Aperture Intensity Mapping (EXCLAIM) is a balloon-borne far-infrared telescope that will survey star formation history over cosmological time scales to improve our understanding of why the star formation rate declined at redshift z < 2, despite continued clustering of dark matter. Specifically,EXCLAIM will map the emission of redshifted carbon monoxide and singly-ionized carbon lines in windows over a redshift range 0 < z < 3.5, following an innovative approach known as intensity mapping. Intensity mapping measures the statistics of brightness fluctuations of cumulative line emissions instead of detecting individual galaxies, thus enabling a blind, complete census of the emitting gas. To detect this emission unambiguously, EXCLAIM will cross-correlate with a spectroscopic galaxy catalog. The EXCLAIM mission uses a cryogenic design to cool the telescope optics to approximately 1.7 K. The telescope features a 90-cm primary mirror to probe spatial scales on the sky from the linear regime up to shot noise-dominated scales. The telescope optical elements couple to six {\mu}-Spec spectrometer modules, operating over a 420-540 GHz frequency band with a spectral resolution of 512 and featuring microwave kinetic inductance detectors. A Radio Frequency System-on-Chip (RFSoC) reads out the detectors in the baseline design. The cryogenic telescope and the sensitive detectors allow EXCLAIM to reach high sensitivity in spectral windows of low emission in the upper atmosphere. Here, an overview of the mission design and development status since the start of the EXCLAIM project in early 2019 is presented.

astro-ph.IM↗