Search NASA⌕ Search

SEARCH · Search NASA

Results for “sample update”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 163 records · Page 9
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Dark Energy Survey Year 3 results: Cosmological constraints from galaxy clustering and weak lensing

We present the first cosmology results from large-scale structure using the full 5000 deg 2 of imaging data from the Dark Energy Survey (DES) Data Release 1. We perform an analysis of large-scale structure combining three two-point correlation functions ( 3 × 2 pt ): (i) cosmic shear using 100 million source galaxies, (ii) galaxy clustering, and (iii) the cross-correlation of source galaxy shear with lens galaxy positions, galaxy–galaxy lensing. To achieve the cosmological precision enabled by these measurements has required updates to nearly every part of the analysis from DES Year 1, including the use of two independent galaxy clustering samples, modeling advances, and several novel improvements in the calibration of gravitational shear and photometric redshift inference. The analysis was performed under strict conditions to mitigate confirmation or observer bias; we describe specific changes made to the lens galaxy sample following unblinding of the results and tests of the robustness of our results to this decision. We model the data within the flat Λ CDM and w CDM cosmological models, marginalizing over 25 nuisance parameters. We find consistent cosmological results between the three two-point correlation functions; their combination yields clustering amplitude S 8 = 0.77 6 - 0.017 + 0.017 and matter density Ω m = 0.33 9 - 0.031 + 0.032 in Λ CDM , mean with 68% confidence limits; S 8 = 0.77 5 - 0.024 + 0.026 , Ω m = 0.35 2 - 0.041 + 0.035 , and dark energy equation-of-state parameter w = - 0.9 8 - 0.20 + 0.32 in w CDM . These constraints correspond to an improvement in signal-to-noise of the DES Year 3 3 × 2 pt data relative to DES Year 1 by a factor of 2.1, about 20% more than expected from the increase in observing area alone. This combination of DES data is consistent with the prediction of the model favored by the Planck 2018 cosmic microwave background (CMB) primary anisotropy data, which is quantified with a probability-to-exceed p = 0.13 –0.48. We find better agreement between DES 3 × 2 pt and Planck than in DES Y1, despite the significantly improved precision of both. When combining DES 3 × 2 pt data with available baryon acoustic oscillation, redshift-space distortion, and type Ia supernovae data, we find p = 0.34 . Combining all of these datasets with Planck CMB lensing yields joint parameter constraints of S 8 = 0.81 2 - 0.008 + 0.008 , Ω m = 0.30 6 - 0.005 + 0.004 , h = 0.68 0 - 0.003 + 0.004 , and ∑ m ν < 0.13 eV (95% C.L.) in Λ CDM ; S 8 = 0.81 2 - 0.008 + 0.008 , Ω m = 0.30 2 - 0.006 + 0.006 , h = 0.68 7 - 0.007 + 0.006 , and w = - 1.03 1 - 0.027 + 0.030 in w CDM .

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Development of Monitoring Techniques for Binderjet Additive Manufacturing of Silicon Carbide Structures

The Transformational Challenge Reactor (TCR) program is leveraging additive manufacturing (AM) technologies to fabricate nuclear components to be assembled into a fully functional microreactor core. Compared with traditional manufacturing technologies, AM technologies allow (1) observation of the manufacturing process at a much higher resolution in real-time using in situ monitoring technologies to capture the sensor signature that scientifically describes each event occurring over time and space and (2) validation of the manufacturing process quality using domain-informed data analytics techniques as a potential qualification and certification methodology for the final component. This report provides an update on the program work on binder jetting in situ process monitoring and associated data analytics results, as well as sample placement and tracking for the subsequent chemical vapor infiltration (CVI) process. Examples are provided to illustrate the progress. Elements of the Digital Thread and data management are discussed in the main document, and an extensive supplemental material section is provided detailing the Digital Platform, as well as its implementation and components. In conclusion the path forward for the next fiscal year is discussed.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2021

ARM Aerosol Measurement Science Group 2019 Strategic Planning Workshop Report

This report summarizes the results of a U.S. Department of Energy Atmospheric Radiation Measurement (ARM) user facility workshop held in November 2019 to advance a science-based strategy for ARM’s aerosol measurement program. This was the second such workshop since the Aerosol Measurement Science Group (AMSG) was chartered in 2015 to enhance coordination of ARM observations of aerosols and atmospheric trace gases with the needs of ARM users. The results presented here reflect the AMSG’s focus in recent years on science-based strategies that will contribute to the increased use of ARM data to fulfill its mission of improving process representations and predictability in climate models. Sessions held during the workshop range from interfacing with models through aerosol sampling strategies to calibration protocols and data products. The strategies set forth here were also developed to be directly relevant to ARM’s updated Decadal Vision. The AMSG workshops have been designed to recommend actions that will enable ARM to evolve and continue to meet its mission. To that end, the AMSG will develop an actionable plan from the recommendations outlined here. Some are well defined and can reasonably be accomplished in the short term. Others are less definite or of a larger scope that calls for a longer-term implementation. Further discussion will be required to develop and prioritize actionable items related to such areas. Task teams comprising the appropriate expertise and perspective from the AMSG and other members of the community will be formed to achieve this outcome. Some particular topics are recognized as high priority, so plans are underway to develop task teams and to hold follow-on discussions to address them. Four areas currently being considered for short, focused discussion are 1) aerosol measurements on the North Slope of Alaska, 2) improving data usability for modeling, 3) strategies for advancing remote sensing, vertical profiling, and distributed measurements of aerosols, and 4) aerosol sampling strategies at existing ARM sites to provide intensive modes of operation to promote data usage for process and modeling studies.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Stochastic modeling and statistical calibration with model error and scarce data

This paper introduces a procedure to assess the predictive accuracy of stochastic models subject to model error and sparse data. Model error is introduced as uncertainty on the coefficients of appropriate polynomial chaos expansions (PCE). The error associated with finite sample size allows us to conceive of these coefficients as statistics of the data that we describe as random variables whose influence on output quantities of interest is evaluated through the extended polynomial chaos expansion (EPCE). A Bayesian data assimilation scheme is introduced to update these expansions by considering the resulting nested chaos expansion as a hierarchical probabilistic model. Stochastic models of quantities of interest (QoI) are thus constructed and efficiently evaluated. Here, the Metropolis–Hastings Markov chain Monte Carlo procedure is used to sample the posterior. Two illustrative analytical and numerical problems are used to demonstrate the proposed approach.

Bayesian inference↗
DOE OSTI2025

Streaming Compression of Scientific Data via Weak-SINDy

Here, in this paper, a streaming weak-SINDy algorithm is developed specifically for compressing streaming scientific data. The production of scientific data, either via simulation or experiments, is undergoing a stage of exponential growth, which makes data compression important and often necessary for storing and utilizing large scientific data sets. As opposed to classical “offline” compression algorithms that perform compression on a readily available data set, streaming compression algorithms compress data “online” while the data generated from simulation or experiments is still flowing through the system. This feature makes streaming compression algorithms well suited for scientific data compression, where storing the full data set offline is often infeasible. This work proposes a new streaming compression algorithm, streaming weak-SINDy, which takes advantage of the underlying data characteristics during compression. The streaming weak-SINDy algorithm constructs feature matrices and target vectors in the online stage via a streaming integration method in a memory efficient manner. The feature matrices and target vectors are then used in the offline stage to build a model through a regression process that aims to recover equations that govern the evolution of the data. For compressing high-dimensional streaming data, we adopt a streaming proper orthogonal decomposition (POD) process to reduce the data dimension and then use the streaming weak-SINDy algorithm to compress the temporal data of the POD expansion. We propose modifications to the streaming weak-SINDy algorithm to accommodate the dynamically updated POD basis. By combining the built model from the streaming weak-SINDy algorithm and a small amount of data samples, the full data flow could be reconstructed accurately at a low memory cost, as shown in the numerical tests.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Deciphering Degradation: Machine Learning on Real-World Performance Data (Final Report)

This project addresses a fundamental flaw in solar PV research and solar project financing; the assumed rate of degradation for solar plants. The solar industry currently relies on an out-dated report that observed a 0.5% degradation rate based on a small sample size of systems (~100). While the research conducted at the time was new and innovative, the solar community has not updated this research and universally applies this 0.5% degradation assumption in financial models. Our project updates this assumption by analyzing observed degradation from the industry’s largest dataset of operating solar assets (>10,000 systems) and creating the first machine-learning model based on these observed results to quantify and identify features that drive degradation. There are two strategic goals for this award: reduce the cost of capital (enable solar to attract more capital) and improve the reliability of solar itself. These dual goals are achieved by leveraging an industry dataset to observe system degradation on a large scale, deploying advanced data analysis and machine learning methods to quantify and predict system reliability, and engaging with industry stakeholders to help them accurately price degradation in financial models.

14 SOLAR ENERGY↗
DOE OSTI2020

Automated Data Review of Analytical Laboratory Results at Los Alamos National Laboratory - 20299

Newport News Nuclear BWXT-Los Alamos, LLC (N3B) collects samples in support of the U.S. Department of Energy's (DOE) Office of Environmental Management (EM) Los Alamos Legacy Cleanup Contract (LLCC). N3B receives and reviews over 1.6 million sample data points annually in support of various ongoing environmental monitoring and remediation projects of the LLCC. N3B must demonstrate and document that reported external analytical laboratory data produced for the LLCC are of sufficient quality to fulfill their intended purpose and to support defensible decision making as described in EPA QA/G4 Guidance for the Data Quality Objectives Process 1994. In 2018, N3B assumed management of the LLCC along with the Environmental Information Management (EIM) database that contains all historical and current environmental data associated with the LLCC. The entire EIM database is shared between N3B, Triad National Security, LLC (Triad), and New Mexico Environment Department (NMED). These three parties jointly manage the database, its configuration, and changes / updates. All environmental data that are entered into EIM are updated and available, on a daily basis, in the linked public database Intellus New Mexico (Intellus). The quality and defensibility of the environmental data generated from sampling activities is a key component of an effective remediation process. Providing quality data is accomplished through a data assessment process that includes examination, verification, and validation. Examination is the assessment of completeness of the deliverables, identification of any reporting errors, and determining the usability of the data based on the laboratory's evaluation of its data as described in the case narrative received with the data. Verification consists of an evaluation of the Electronic Data Deliverables (EDD) data report to determine the extent to which the external analytical laboratories met method and contract-specific quality control and reporting requirements. Validation consists of determining the data quality and the extent to which the external analytical laboratories accurately and completely reported all sample and quality control results and satisfied all contract requirements. EIM contains an automatic Data Validation Module which performs automated data review (DVM ADR). DVM ADR is a tool to assist in the validation process. When DVM ADR is used in conjunction with manual examination of sample data packages, the combination of the two will meet and exceed the requirements of verification. N3B recognized an opportunity for process improvement, focusing on DVM ADR configuration and enhancements in EIM. Testing EIM's configuration provided proof of the DVM ADR's capabilities and flexibility to accurately perform routine data checks based on analytical methods and regulatory requirements. In addition, the DVM ADR module was improved through enhancements for all analytes, particularly upgrades for radiochemistry data. Extensive testing of the DVM ADR module occurred using EDDs from actual laboratory analyses on the EIM testing site. During this process, N3B manipulated EDD information to verify that the actual outcomes matched the expected outcomes. The results of this testing were shared with the database architects, and configuration improvements were identified to address these results. During this process, N3B identified that the radiochemical DVM ADR capabilities were underutilized, and so enhanced the DVM ADR functionality with respect to radioanalytical assessment. N3B environmental data uploads to Intellus on a daily basis from EIM, once the analytical data undergoes examination and verification. As such, it is important to have a high level of confidence in the quality and defensibility of the data. The process of manual examination, along with the DVM ADR, in conjunction with full validation of a percentage the data specified through the Data Quality Objectives greatly increases efficiency of data review and confidence level of the quality of the data, and gives the project managers, governmental offices, and the public expedited access to high-quality data. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2021

A parallel evolutionary multiple-try metropolis Markov chain Monte Carlo algorithm for sampling spatial partitions

We develop an Evolutionary Markov Chain Monte Carlo (EMCMC) algorithm for sampling spatial partitions that lie within a large, complex, and constrained spatial state space. Our algorithm combines the advantages of evolutionary algorithms (EAs) as optimization heuristics for state space traversal and the theoretical convergence properties of Markov Chain Monte Carlo algorithms for sampling from unknown distributions. Local optimality information that is identified via a directed search by our optimization heuristic is used to adaptively update a Markov chain in a promising direction within the framework of a Multiple-Try Metropolis Markov Chain model that incorporates a generalized Metropolis-Hastings ratio. We further expand the reach of our EMCMC algorithm by harnessing the computational power afforded by massively parallel computing architecture through the integration of a parallel EA framework that guides Markov chains running in parallel.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

FUSARIUM-ID v.3.0: An Updated, Downloadable Resource for Fusarium Species Identification

Species within Fusarium are of global agricultural, medical, and food/feed safety concern and have been extensively characterized. However, accurate identification of species is challenging and usually requires DNA sequence data. FUSARIUM-ID ( http://isolate.fusariumdb.org/blast.php ) is a publicly available database designed to support the identification of Fusarium species using sequences of multiple phylogenetically informative loci, especially the highly informative ~680-bp 5' portion of the translation elongation factor 1-alpha (TEF1) gene that has been adopted as the primary barcoding locus in the genus. However, FUSARIUM-ID v.1.0 and 2.0 had several limitations, including inconsistent metadata annotation for the archived sequences and poor representation of some species complexes and marker loci. Here, we present FUSARIUM-ID v.3.0, which provides the following improvements: (i) additional and updated annotation of metadata for isolates associated with each sequence, (ii) expanded taxon representation in the TEF1 sequence database, (iii) availability of the sequence database as a downloadable file to enable local BLAST queries, and (iv) a tutorial file for users to perform local BLAST searches using either freely available software, such as SequenceServer, BLAST+ executable in the command line, and Galaxy, or the proprietary Geneious software. FUSARIUM-ID will be updated on a regular basis by archiving sequences of TEF1 and other loci from newly identified species and greater in-depth sampling of currently recognized species.

Plant Sciences↗
DOE OSTI2023

Updated Application of Frequency of Detection Methods for the INL Site Ambient Air Monitoring Network

This report presents a quantitative assessment of the current INL Site air monitoring network using frequency of detection (FD) methods. The first assessment of the INL network was performed in 2015 and made recommendations for improving the network. As a result of changes made in response to the recommendations and the addition of new source locations, the network was modified and reassessed in 2017. Since 2017, administration of the air sampling program has been consolidated under one contractor, which resulted in additional changes to the network (sampler numbers and locations) and changes in radionuclide detection levels. As a result of these changes and others, an updated assessment of the INL Site ambient air monitoring network was performed. The same two exposure scenarios used in previous assessments were used for this assessment: a resident scenario and a shepherd/rancher scenario. The resident was assumed to be continuously present at their residence/business/farm operation outside the INL Site boundary while the shepherd/rancher was assumed to be present 24-hours at the nearest INL grazing allotment boundary in each of the 22.5-degree sectors along the sector centerline from each source. Updates to both the resident and shepherd/rancher receptor locations were included. Other changes include updates to flow rates for stack sources, expansion of the list of important radionuclides based on the most recent National Emission Standards for Hazardous Pollutants (NESHAPs) analysis, and updated dose coefficients. The assessment was conducted to determine whether the current INL monitoring network is capable of detecting releases of important radionuclides from INL Site sources that have the potential to exceed a conservative dose threshold for the two exposure scenarios. The assessment revealed that for the resident scenario, the current network meets the desired performance objective (FD = 95%) for all radionuclides and sources except for Cl-36 from the TRA-770 stack (94.4%). For the shepherd/rancher scenario, the FD performance objective is met for all radionuclides and sources except tritium from MFC-774 and TAN 679 (91% for both). An investigation of reported emissions for the past three years revealed that Cl 36 is not emitted from TRA-770, and routine tritium emissions from MFC-774 and TAN-679 are very small and the sources are likely incapable of emitting enough tritium to cause a release that should be detectable by the monitoring network. This assessment is based on a conservative dose threshold. This coupled with fact that the FD for Cl 36 is only slightly less than the performance objective and Cl-36 is not emitted from TRA-770, modifying the network (i.e. adding another sampler, increasing sampler flow rate, moving samplers) to meet the 95% performance objective for this radionuclide/source/receptor scenario is not warranted. Similarly, because tritium emissions from MFC-774 and TAN-679 are very small and these two sources are likely incapable of causing a dose due to tritium release that should be detectable by the network, modifications to increase tritium detection for these sources is also unwarranted at this time. However, if it is required to meet the performance objective for tritium for all sources and receptor scenarios, additional analysis determined the FD could be raised from 91% to > 99% for both sources by adding two tritium samplers to the network.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2020

Play Fairway analysis of geothermal resources across the State of Hawai‘i: 4. Updates with new groundwater chemistry, subsurface stress analysis, and focused geophysical surveys

This paper is the fourth in a series on a play fairway analysis of geothermal resources across the State of Hawai‘i. Here we describe recent exploration activities that include groundwater sampling in ten locations statewide, as well as geophysical surveys on Lana‘i, across the SW rift of Haleakala Volcano (Maui), and surrounding Mauna Kea Volcano (Hawai‘i Island). We derive a first-order method for computing topographic stress using Green’s functions, finding that topography can induce appreciable crustal stress. We develop a new method for incorporating depth information about resistivity, density, and topographic stresses into our previously published equations to calculate resource probability and confidence. We incorporate newly collected and modeled data into our calculations to update statewide maps of probability and confidence. Lana‘i Island and southeast Mauna Kea are identified as the top targets for exploratory drilling. The east rift of Haleakala Volcano, the southern-most region of Mauna Loa’s SW rift (Hawai‘i Island), and central Kauai are identified as targets for geophysical surveying.

15 GEOTHERMAL ENERGY↗
DOE OSTI2024

Efficient distributed continual learning for steering experiments in real-time

Deep learning has emerged as a powerful method for extracting valuable information from large volumes of data. However, when new training data arrives continuously (i.e., is not fully available from the beginning), incremental training suffers from catastrophic forgetting (i.e., new patterns are reinforced at the expense of previously acquired knowledge). Training from scratch each time new training data becomes available would result in extremely long training times and massive data accumulation. Rehearsal-based continual learning has shown promise for addressing the catastrophic forgetting challenge, but research to date has not addressed performance and scalability. To fill this gap, we propose an approach based on a distributed rehearsal buffer that efficiently complements data-parallel training on multiple GPUs to achieve high accuracy, short runtime, and scalability. It leverages a set of buffers (local to each GPU) and uses several asynchronous techniques for updating these local buffers in an embarrassingly parallel fashion, all while handling the communication overheads necessary to augment input minibatches using unbiased, global sampling. We further propose a generalization of rehearsal buffers to support both classification and generative learning tasks, as well as more advanced rehearsal strategies (notably Dark Experience Replay, leveraging knowledge distillation). We illustrate this approach with a real-life HPC streaming application from the domain of ptychographic image reconstruction. Furthermore, we run extensive experiments on up to 128 GPUs of the ThetaGPU supercomputer to compare our approach with baselines representative of training-from-scratch (the upper bound in terms of accuracy) and incremental training (the lower bound). Results show that rehearsal-based continual learning achieves a top-5 validation accuracy close to the upper bound, while simultaneously exhibiting a runtime close to the lower bound.

Asynchronous data management↗
DOE OSTI2026

2008 East Tennessee Household Travel Survey

The 2008 East Tennessee Household Travel Survey, conducted by NuStats, was a comprehensive study of travel behavior in Knox, Blount, Anderson, Jefferson, Loudon, Roane, Sevier, and Union counties. The purpose of the survey was to collect weekday travel characteristics of households in the eight-county region during a 24-hour timeframe. The data will be used by the Knoxville Regional Transportation Planning Organization and local agencies to update transportation and air quality models and to identify transportation needs in the region. The survey collected demographic and travel data from a sample of 1,400 households between February and May 2008. In total, 12,012 trips were reported by 3,301 persons.

1Hz data↗
DOE OSTI2026

Meteoric 10Be Flux Calibration Data for the East River Watershed, Colorado, USA

This data package contains tabular and geospatial data used to quantify and model meteoric beryllium-10 fluxes in the East River watershed, Colorado, USA. The tabular component includes calibration-site data from five glacial moraine sites and includes environmental variables used to evaluate spatial controls on meteoric 10Be delivery, including elevation, mean annual precipitation (MAP), mean snow depth, and mean snow water equivalent (SWE). These site-level data were used to compare observed fluxes with environmental gradients across the watershed and to evaluate the effects of erosion correction on flux estimates. The package also includes supporting slope and curvature values used to assess topographic inputs to the erosion analysis. A second component of the data package contains updated manuscript tables and regression outputs used to summarize the relationships between meteoric 10Be flux and environmental predictors. These tables include meteoric 10Be sample information and AMS results, site-level environmental values, site-level meteoric 10Be inventory and flux values, watershed-averaged predicted fluxes, soil bulk density measurements, fine-fraction values, soil pH measurements, and regression statistics including slope, intercept, coefficient of determination, and p-value. The regression products include both standard linear regressions and regressions in which the intercept is constrained to pass through zero, and they support the analyses presented in the companion manuscript. Together, these tabular files provide the numerical basis for the manuscript tables and the regression-based interpretation of meteoric 10Be flux variability in a snow-dominated mountain watershed. The geospatial component of the package consists of GeoTIFF raster files used to generate the map products presented in Figures 2 and 6 of the companion manuscript. These rasters represent watershed-scale spatial layers for environmental variables and regression-based predictions of meteoric 10Be flux. This dataset contains comma-separated values files (.csv), Microsoft Excel files (.xlsx), GeoTIFF raster files (.tif), and upporting metadata files, including CSV data dictionaries and readme text files (.csv, .txt). The tabular files can be opened with standard spreadsheet software, and the raster files can be viewed and analyzed in GIS software such as ArcGIS Pro or QGIS. Together, these files document the numerical and spatial datasets used to calibrate and predict meteoric 10Be delivery in the East River watershed.

East River↗
DOE OSTI2022

The Fourth Catalog of Active Galactic Nuclei Detected by the Fermi Large Area Telescope: Data Release 3

Abstract An incremental version of the fourth catalog of active galactic nuclei (AGNs) detected by the Fermi Large Area Telescope is presented. This version (4LAC-DR3) derives from the third data release of the 4FGL catalog based on 12 yr of E > 50 MeV gamma-ray data, where the spectral parameters, spectral energy distributions (SEDs), yearly light curves, and associations have been updated for all sources. The new reported AGNs include 587 blazar candidates and four radio galaxies. We describe the properties of the new sample and outline changes affecting the previously published one. We also introduce two new parameters in this release, namely the peak energy of the SED high-energy component and the corresponding flux. These parameters allow an assessment of the Compton dominance, the ratio of the inverse-Compton to the synchrotron-peak luminosities, without relying on X-ray data.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning

Large language models (LLMs) are increasingly adapted to downstream tasks via reinforcement learning (RL) methods like Group Relative Policy Optimization (GRPO), which often require thousands of rollouts to learn new tasks. We argue that the interpretable nature of language often provides a much richer learning medium for LLMs, compared to policy gradients derived from sparse, scalar rewards. To test this, we introduce GEPA (Genetic-Pareto), a prompt optimizer that thoroughly incorporates natural language reflection to learn high-level rules from trial and error. Given any AI system containing one or more LLM prompts, GEPA samples trajectories (e.g., reasoning, tool calls, and tool outputs) and reflects on them in natural language to diagnose problems, propose and test prompt updates, and combine complementary lessons from the Pareto frontier of its own attempts. As a result of GEPA's design, it can often turn even just a few rollouts into a large quality gain. Across six tasks, GEPA outperforms GRPO by 6% on average and by up to 20%, while using up to 35x fewer rollouts. GEPA also outperforms the leading prompt optimizer, MIPROv2, by over 10% (e.g., +12% accuracy on AIME-2025), and demonstrates promising results as an inference-time search strategy for code optimization. We release our code at https://github.com/gepa-ai/gepa.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

BioXTAS RAW 2 : new developments for a free open-source program for small-angle scattering data reduction and analysis

BioXTAS RAW is a free open-source program for reduction, analysis and modelling of biological small-angle scattering data. Here, the new developments in RAW version 2 are described. These include improved data reduction using pyFAI ; updated automated Guinier fitting and D max finding algorithms; automated series ( e.g. size-exclusion chromatography coupled small-angle X-ray scattering or SEC-SAXS) buffer- and sample-region finding algorithms; linear and integral baseline correction for series; deconvolution of series data using regularized alternating least squares ( REGALS ); creation of electron-density reconstructions using electron density via solution scattering ( DENSS ); a comparison window showing residuals, ratios and statistical comparisons between profiles; and generation of PDF reports with summary plots and tables for all analysis. Furthermore, there is now a RAW API, which can be used without the graphical user interface (GUI), providing full access to all of the functionality found in the GUI. In addition to these new capabilities, RAW has undergone significant technical updates, such as adding Python 3 compatibility, and has entirely new documentation available both online and in the program.

97 MATHEMATICS AND COMPUTING↗