Pareto Prompt Optimization
Explore the source record for details and available documents.
97 MATHEMATICS AND COMPUTING↗
SEARCH · Engineering Papers
Results for “prompt optimization”
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.
Prompt Phrase Ordering Using Large Language Models in HPC: Evaluating Prompt Sensitivity
Large language models (LLMs) have demonstrated effective performance in domain-specific tasks, often requiring a well-designed prompt to guide their responses. However, optimizing the right prompt is challenging due to prompt sensitivity—the phenomenon where small changes in the prompt can lead to significant variations in performance. In this study, we evaluate prompt performance by examining all permutations of independent phrases to investigate prompt sensitivity and robustness. We used two datasets: the GSM8k dataset, which assesses mathematical reasoning, and a custom template prompt for summarizing database metadata. Our goal was to evaluate the performance across all permutations of a sequence of prompt phrases. The study was conducted using the llama3-instruct- 7B model hosted on Ollama, with computations parallelized in a high-performance computing environment. By comparing the average index of phrases in the best and worst-performing prompts, we found that the order of independent phrases within a prompt significantly impacts LLM performance. Additionally, we used Hamming distance to assess changes between phrase orderings, concluding that prompt modifications can dramatically affect scores, often by almost random chance. These findings support existing research on prompt sensitivity. We discuss the challenges of prompt optimization, noting that altering phrases in a successful prompt does not always result in another successful prompt.
97 MATHEMATICS AND COMPUTING↗
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↗
Pushing the Limits of Pulse Shape Discrimination in a Large Liquid Xenon Detector
The LUX-ZEPLIN (LZ) experiment is a direct-detection dark matter experiment, optimized to search for weakly interacting massive particles (WIMPs) through WIMP-nucleon interactions. The main challenge in dark matter detection is differentiating between WIMP signals and background events. In LZ, the ratio of ionization to scintillation signals (charge-to-light) is the primary method for rejecting electronic recoil (ER) background. Pulse shape discrimination (PSD) offers a method for additional ER backgrounds rejection in liquid xenon detectors. In this paper, the discrimination power of PSD with the LZ experiment is discussed. To precisely characterize the scintillation pulse shape, an analysis framework is developed to reconstruct the detection time of individual photons. Using LZ calibration data, the photon-timing prompt fraction discriminator is optimized and achieves ER leakage as low as $15\%$. For specific background processes such as $^{124}$Xe double electron capture, the leakage is reduced further to about $5\%$. PSD is combined with charge-to-light to form two-factor discrimination (TFD). The optimized TFD performance is compared with the performance of the charge-to-light method, with the corresponding false positive rate reduced by up to a factor of two for large scintillation pulses. Finally, PSD and TFD are applied to data from LZ's WS2024 run and their performance is summarized.
Akerib, D. S. [SLAC; KIPAC, Menlo Park]↗
Nanoscopic strain evolution in single-crystal battery positive electrodes
Single-crystal Ni-rich layered oxides (SC-NMC) with a grain-boundary-free configuration have effectively addressed the long-standing cracking issue of conventional polycrystalline Ni-rich materials (PC-NMC) in lithium-ion batteries, prompting a shift in optimization strategies. However, continued reliance on anisotropic lattice volume change—a well-established failure indicator in PC-NMC—as a metric for understanding strain and guiding compositional design for SC-NMC becomes controversial. Here, in this study, by leveraging multiscale diagnostic techniques, we unravelled the distinct nanoscopic strain evolution in SC-NMC during battery operation, challenging the conventional composition-driven strategies and mechanical degradation indicators used for PC-NMC. Through particle-level chemomechanical analysis, we reveal a decoupling between mechanical stability and lattice volume change in SC-NMC, identifying that structural instability in SC materials is primarily driven by multidimensional lattice distortions induced by kinetics-driven reaction heterogeneity and progressively deactivating chemical phases. Using this mechanical failure mode, we redefine the roles of cobalt and manganese in maintaining mechanical stability. Unlike cobalt’s detrimental role in PC-NMC, we find cobalt to be critical in enhancing the longevity of SC-NMC by mitigating localized strain along the extended diffusion pathway, whereas manganese exacerbates mechanical degradation.
36 MATERIALS SCIENCE↗
Compression Molding of Composite of Recycled HDPE and Recycled Tire Particles
Plastic and rubber recycling is an effective means of reducing solid waste to the environment and preserving natural resources. A project aimed at developing a new composite material from recycled high density polyethylene (HDPE) and recycled rubber is currently being conducted at Eastern Illinois University. The recycled plastic pellets with recycled rubber particles are extruded into some HDPE/rubber composite strands. The strand can be further cut into pellets that can be used to fabricate other material forms or products. This experiment was inspired by the above-mentioned research activity. In order to measure Durometer hardness of the extruded composite, a specimen with relatively large dimensions was needed. Thus, compression molding was used to form a cylindrical specimen of 1 in. diameter and 1 in. thickness. The initial poor quality of the molded specimen prompted a need to optimize the processing parameters such as temperature, holding time, and pressure. Design of experiment (DOE) was used to obtain optimum combination of the parameters.
Liu, Ping↗
Machine Learning for Scalable and Optimal Load Shedding Under Power System Contingency
Prompt and effective corrective actions in response to unexpected contingencies are crucial for improving power system resilience and preventing cascading blackouts. The optimal load shedding (OLS) accounting for network limits has the potential to address the diverse system-wide impacts of contingency scenarios as compared to traditional local schemes. However, due to the fast cascading propagation of initial contingencies, real-time OLS solutions are challenging to attain in large systems with high computation and communication needs. In this paper, we propose a decentralized design that leverages offline training of a neural network (NN) model for individual load centers to autonomously construct the OLS solutions from locally available measurements. Our learning-for-OLS approach can greatly reduce the computation and communication needs during online emergency responses, thus preventing the cascading propagation of contingencies for enhanced power grid resilience. Numerical studies on both the IEEE 118-bus system and a synthetic Texas 2000-bus system have demonstrated the efficiency and effectiveness of our scalable OLS learning design for timely power system emergency operations.
24 POWER TRANSMISSION AND DISTRIBUTION↗
Visual Instance-aware Prompt Tuning
Visual Prompt Tuning (VPT) has emerged as a parameter-efficient fine-tuning paradigm for vision transformers, with conventional approaches utilizing dataset-level prompts that remain the same across all input instances. We observe that this strategy results in sub-optimal performance due to high variance in downstream datasets. To address this challenge, we propose Visual Instance-aware Prompt Tuning (ViaPT), which generates instance-aware prompts based on each individual input and fuses them with dataset-level prompts, leveraging Principal Component Analysis (PCA) to retain important prompting information. Moreover, we reveal that VPT-Deep and VPT-Shallow represent two corner cases based on a conceptual understanding, in which they fail to effectively capture instance-specific information, while random dimension reduction on prompts only yields performance between the two extremes. Instead, ViaPT overcomes these limitations by balancing dataset-level and instance-level knowledge, while reducing the amount of learnable parameters compared to VPT-Deep. Extensive experiments across 34 diverse datasets demonstrate that our method consistently outperforms state-of-the-art baselines, establishing a new paradigm for analyzing and optimizing visual prompts for vision transformers.
Xiao, Xi [ORNL] (ORCID:0009000009316982)↗
Comparison and Application of Battery Modeling Methods for Conceptual Electric Aircraft
Growing interest in design and optimization of electrified aircraft propulsion concepts prompts the need for accurate, flexible, and efficient methods to model battery systems. Presented in this paper are three battery modeling methods that have been used at NASA’s Glenn Research Center, each representing different mathematical or electrical approaches. Thévenin equivalent circuit, normalization, and curve-fitting methods are compared against battery cell test data for the X-57 Maxwell electric aircraft technology demonstrator. The methods are then applied in a simple multidisciplinary optimization context using NASA’s Six-Passenger Electric Quadrotor concept to determine their applicability and performance. The normalization method achieves the highest accuracy for steady and unsteady discharge rates with a voltage mean error percentage of 0.423% and 1.186%, respectively. Optimal quadrotor mission range between the models varies up to 0.5 nmi, identifying current battery modeling methods as a potentially significant contributor to mission analysis error. A set of relevant tools and techniques for conceptual battery modeling are identified in this paper, with conclusions made on the utility of each modeling approach for various design challenges.
power↗
Terminal-area flight experience with the NASA Terminal Configured Vehicle
Increases in the volume of air traffic around major terminals, as well as the advent of MLS and other sophisticated ATC techniques, have prompted the development of aircraft specifically optimized for flight in terminal areas. A modified Boeing 737, developed by NASA, is presented as an illustration of a Terminal Configured Vehicle (TCV). Among the TCV systems discussed are the electronic attitude direction indicator (EADI) and the electronic horizontal situation indicator (EHSI), both of which are advanced display systems capable of positioning the runway center line from a variety of glide slope paths, and with reference to time, i.e., the on-board computer can predict where the aircraft will be at a given number of seconds from the time of the analysis. Thus, 4-D navigation (time path) can be selected as a control mode, together with vertical and horizontal path modes.
Morello, S. A.↗
Joint ASRS and NASA Callback on FANS-1 Datalink Operational Incidents
The recent certification and implementation of the Future Air Navigation System (FANS- 1) was based on the benefits of reduced separation minima, ability to optimize flight plans enroute, and the prompt transmission and reception of messages between oceanic controllers and 747-400 aircraft transiting the Pacific. In addition, FANS was intended to supersede High Frequency (HF) radio which has been the staple, yet problematic, long distance communication link for years. However, in the three initial years of operation, FANS has revealed its own unique operational issues. Although some technical and engineering problems have been studied and addressed, little research has been conducted on human factors issues associated with the use of FANS on the flight deck. This lack of prior data on the operational use of FANS prompted a joint NASA/Aviation Safety Reporting System (ASRS) inquiry into FANS datalink. In addition, two foreign air carriers and their respective safety agencies were recruited to participate. This international FANS research effort consisting of three aviation safety agencies and three international commercial air-carriers has provided a unique opportunity for conducting human factors research in an operational environment. An ASRS 'callback' format was chosen for the study as a practical and viable method for capturing FANS events on the flight deck. Initially, 747-400 pilots were encouraged to submit reports to their respective safety agencies if they had experienced any positive or negative incidents with the use of FANS. Upon receipt of a report, it was evaluated and if deemed pertinent to the study, a telephone interview or 'callback' was conducted on the FANS incident to elicit further details, capture the key events, and gather contextual information. Once the operational data collection phase was completed, the reports and interviews were analyzed with two purposes in mind: primarily to identify and address problematic human factors issues with current FANS technology, and, secondarily, to provide 'lessons learned' for future datalink systems and implementation.
Moses, John↗
Bayesian Optimization of Catalysis with In-Context Learning
Large language models (LLMs) can perform accurate classification with zero or few examples through in-context learning (ICL), allowing the model to observe query-relevant examples at inference time and eliminating the need for additional weight updates to generalize beyond its original training data. We extend this capability to regression with uncertainty estimation using frozen LLMs (e.g., GPT-4o, Gemini), enabling Bayesian optimization (BO) in natural language without explicit model training or feature engineering. We apply this to materials discovery by representing materials as synthesis and testing procedures for use in natural language prompts. This Bayesian, design-first approach prioritizes optimization toward target material properties before detailed characterization, in contrast to conventional experimental workflows that often emphasize characterization of suboptimal materials. On benchmarks like aqueous solubility and oxidative coupling of methane (OCM), BO-ICL matches or outperforms Gaussian processes. In live experiments on the reverse water–gas shift (RWGS) reaction, BO-ICL identifies multimetallic catalysts that approach equilibrium CO yield within 6 and 10 iterations from a pool of 3,700 and 360,000 candidates, respectively. Our method redefines materials representation and accelerates discovery, with broad applications across catalysis, materials science, and AI.
Calibration↗
SmileyLlama: modifying large language models for directed chemical space exploration
Here we show that large language models (LLMs) can be transformed via supervised fine-tuning of engineered prompts into SmileyLlama for exploring the chemical space of drug molecules. We benchmark SmileyLlama against pretrained LLMs and chemical language models trained from scratch for generating valid and novel drug-like molecules, and use direct preference optimization to both improve SmileyLlama’s adherence to a prompt and as part of the iMiner reinforcement learning framework to predict molecules with optimized three-dimensional conformations and high binding affinity to drug targets. By training an LLM to speak directly as a chemical language model, while retaining most of its natural language capabilities, we show that SmileyLlama can reliably generate molecules with user-specified properties rather than acting only as a chatbot with knowledge of chemistry or as a virtual assistant. While SmileyLlama is geared toward drug discovery, the supervised fine-tuning/direct preference optimization/LLM framework can be extended to other chemical, biological and materials applications.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Advanced cargo aircraft may offer a potential renaissance in freight transportation
The increasing demand for air freight transportation has prompted studies of large, aerodynamically efficient cargo-optimized aircraft capable of carrying intermodal containers, which are typically 8 x 8 x 20 ft. Studies have accordingly been conducted within NASA to ascertain the specifications and projected operating costs of such a vehicle, as well as to identify critical, development-pacing technologies. Attention is here given not only to the rather conventional, 10-turbofan engined configuration thus arrived at, but numerous innovative configurations featuring such concepts as spanloading, removable cargo pods, and ground effect.
Morris, Shelby J.↗
Martian Radiative Transfer Modeling Using the Optimal Spectral Sampling Method
The large volume of existing and planned infrared observations of Mars have prompted the development of a new martian radiative transfer model that could be used in the retrievals of atmospheric and surface properties. The model is based on the Optimal Spectral Sampling (OSS) method [1]. The method is a fast and accurate monochromatic technique applicable to a wide range of remote sensing platforms (from microwave to UV) and was originally developed for the real-time processing of infrared and microwave data acquired by instruments aboard the satellites forming part of the next-generation global weather satellite system NPOESS (National Polarorbiting Operational Satellite System) [2]. As part of our on-going research related to the radiative properties of the martian polar caps, we have begun the development of a martian OSS model with the goal of using it to perform self-consistent atmospheric corrections necessary to retrieve caps emissivity from the Thermal Emission Spectrometer (TES) spectra. While the caps will provide the initial focus area for applying the new model, it is hoped that the model will be of interest to the wider Mars remote sensing community.
Eluszkiewicz, J.↗
Data Analysis for the LISA Pathfinder Mission
The LTP (LISA Technology Package) is the core part of the Laser Interferometer Space Antenna (LISA) Pathfinder mission. The main goal of the mission is to study the sources of any disturbances that perturb the motion of the freely-falling test masses from their geodesic trajectories as well as 10 test various technologies needed for LISA. The LTP experiment is designed as a sequence of experimental runs in which the performance of the instrument is studied and characterized under different operating conditions. In order to best optimize subsequent experimental runs, each run must be promptly analysed to ensure that the following ones make best use of the available knowledge of the instrument ' In order to do this, all analyses must be designed and tested in advance of the mission and have sufficient built-in flexibility to account for unexpected results or behaviour. To support this activity, a robust and flexible data analysis software package is also required. This poster presents two of the main components that make up the data analysis effort: the data analysis software and the mock-data challenges used to validate analysis procedures and experiment designs.
Thorpe, James Ira↗
Simulation of the radiological impact during selected space travel scenarios using the Monte Carlo code FLUKA
Radiation is one of the major challenges of space exploration and can negatively impact both biological and electronic systems, particularly in the case of long-term journeys or if the spaceship features inadequate shielding. Here, in this work, the cumulative dose levels from prompt radiation in the spacecraft are quantified alongside the residual dose contributions arising from activation of vessel components. The radiological impact was assessed for various space exploration scenarios, considering the same spaceship model featuring three shielding design variants. In each scenario, the radiation environment was generated with the Monte Carlo particle transport and interaction code FLUKA. These results can be used to quantify the contribution of prompt and residual dose in spacefaring ventures and help determine optimal radiation shielding needed to mitigate the overall radiological impact on both astronauts and equipment.
73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
Alternate charging profiles for the onboard nickel cadmium batteries of the Explorer Platform/Extreme Ultraviolet Explorer
The Explorer Platform/Extreme Ultraviolet Explorer (EP/EUVE) spacecraft power is provided by the Modular Power Subsystems (MPS) which contains three 50 ampere-hour Nickel Cadmium (NiCd) batteries. The batteries were fabricated by McDonnell Douglas Electronics Systems Company, with the cells fabricated by Gates Aerospace Batteries (GAB), Gainesville, Florida. Shortly following launch, the battery performance characteristics showed similar signatures as the anomalous performance observed on both the Upper Atmosphere Research Satellite (UARS) and the Compton Gamma Ray Observatory (CGRO). This prompted the development and implementation of alternate charging profiles to optimize the spacecraft battery performance. The Flight Operations Team (FOT), under the direction of Goddard Space Flight Center's (GSFC) EP/EUVE Project and Space Power Applications Branch have monitored and managed battery performance through control of the battery Charge to Discharge (C/D) ratio and implementation of a Solar Array (SA) offset. This paper provides a brief overview of the EP/EUVE mission, the MPS, the FOT's battery management for achieving the alternate charging profile, and the observed spacecraft battery performance.
Rao, Gopalakrishna M.↗