Karl Pazdernik

Rounak Meyur, Hung Phan, Koby Hayashi et al.

The National Environment Policy Act (NEPA) stands as a foundational piece of environmental legislation in the United States, requiring federal agencies to consider the environmental impacts of their proposed actions. The primary mechanism for achieving this is through the preparation of Environmental Assessments (EAs) and, for significant impacts, comprehensive Environmental Impact Statements (EIS). Large Language Model (LLM)s' effectiveness in specialized domains like NEPA remains untested for adoption in federal decision-making processes. To address this gap, we present NEPA Question and Answering Dataset (NEPAQuAD), the first comprehensive benchmark derived from EIS documents, along with a modular and transparent evaluation pipeline, MAPLE, to assess LLM performance on NEPA-focused regulatory reasoning tasks. Our benchmark leverages actual EIS documents to create diverse question types, ranging from factual to complex problem-solving ones. We built a modular and transparent evaluation pipeline to test both closed- and open-source models in zero-shot or context-driven QA benchmarks. We evaluate five state-of-the-art LLMs using our framework to assess both their prior knowledge and their ability to process NEPA-specific information. The experimental results reveal that all the models consistently achieve their highest performance when provided with the gold passage as context. While comparing the other context-driven approaches for each model, Retrieval Augmented Generation (RAG)-based approaches substantially outperform PDF document contexts, indicating that neither model is well suited for long-context question-answering tasks. Our analysis suggests that NEPA-focused regulatory reasoning tasks pose a significant challenge for LLMs, particularly in terms of understanding the complex semantics and effectively processing the lengthy regulatory documents.

Sai Munikoti, Ian Stewart, Chengping Chai et al.

The application of generalist multimodal models (GMMs) to specialized scientific domains remains limited due to the scarcity of comprehensive domain-specific datasets that integrate multiple data modalities beyond text and images. In seismology, understanding earthquake phenomena requires the synthesis of timeseries waveform data, geographical imagery, and contextual metadata, a multimodal integration absent in existing seismic datasets. We present MultiSeismo, a large scale structured multimodal seismic dataset, comprising over 16K seismic events spanning 13 years (2010 to 2023) across diverse geographical regions. Each event data integrates waveform recordings from global station networks, intensity maps, population exposure visualizations, and a comprehensive textual description within a standardized JSON format. We additionally develop MISCE, a multimodal instruction set on top of raw data to enable supervised training and evaluation of GMMs on seismic reasoning tasks ranging from basic information retrieval to complex cross modal analysis. We leverage MISCE to finetune an existing multimodal model (Unified IO 2) enhanced with a specialized timeseries encoder, which yields SeisModal, the first domain specific multimodal model for comprehensive seismic analysis. Evaluation of state of the art multimodal models on MultiSeismo reveals significant challenges, particularly with time-series data processing for general purpose models, while demonstrating SeisModal's superior performance on seismic multimodal reasoning tasks. These results prove that MultiSeismo provides a rigorous benchmark for future multimodal research in seismology and validate the success of our domain specific architectural adaptations.

Sameera Horawalavithana, Lauren Phillips, Ian Stewart et al.

Vision-Language Models (VLMs) have rapidly advanced by leveraging powerful pre-trained Large Language Models (LLMs) as core reasoning backbones. As new and more capable LLMs emerge with improved reasoning, instruction-following, and generalization, there is a pressing need to efficiently update existing VLMs to incorporate these advancements. However, the integration of new LLMs into VLMs, particularly how the evolving LLMs contribute to multimodal reasoning, alignment, and task-specific performance remains underexplored. Addressing this gap is important for VLM development, given the rapid evolution of pretrained LLM backbones. This study presents a controlled and systematic investigation of how changes in the pretrained LLM backbone affect downstream VLM task performance. By having the vision encoder, training data, and post-training algorithm remain same across LLAMA-1, LLAMA-2, and LLAMA-3 based VLMs, we find that newer LLM backbones do not always lead to better VLMs, but the performance depends on the downstream VLM task. For example, in visual question and answering tasks, newer LLM backbones tend to solve different questions rather than just more questions, and our analysis shows this is driven by differences in how the models process information, including better calibrated confidence and more stable internal representations. We also find that some VLM capabilities appear only in the newest LLM generation, while tasks that depend mainly on visual understanding see little benefit from a newer LLM backbone.

Ian Stewart, Sameera Horawalavithana, Brendan Kennedy et al.

Multimodal foundation models (MFMs) such as OFASys show the potential to unlock analysis of complex data such as images, videos, and audio data via text prompts alone. However, their performance may suffer in the face of text input that differs even slightly from their training distribution, which is surprising considering the use of modality-specific data to "ground" the text input. This study demonstrates that prompt instability is a major concern for MFMs, leading to a consistent drop in performance across all modalities, but that instability can be mitigated with additional training with augmented data. We evaluate several methods for grounded prompt perturbation, where we generate perturbations and filter based on similarity to text and/or modality data. After re-training the models on the augmented data, we find improved accuracy and more stable performance on the perturbed test data regardless of perturbation condition, suggesting that the data augmentation strategy helps the models handle domain shifts more effectively. In error analysis, we find consistent patterns of performance improvement across domains, suggesting that retraining on prompt perturbations tends to help general reasoning capabilities in MFMs.

Alex Hagen, Karl Pazdernik, Nicole LaHaye et al.

Measurement of uncertainty of predictions from machine learning methods is important across scientific domains and applications. We present, to our knowledge, the first such technique that quantifies the uncertainty of predictions from a classifier and accounts for both the classifier's belief and performance. We prove that our method provides an accurate estimate of the probability that the outputs of two neural networks are correct by showing an expected calibration error of less than 0.2% on a binary classifier, and less than 3% on a semantic segmentation network with extreme class imbalance. We empirically show that the uncertainty returned by our method is an accurate measurement of the probability that the classifier's prediction is correct and, therefore has broad utility in uncertainty propagation.

Alex Hagen, Shane Jackson, James Kahn et al.

Statistical testing is widespread and critical for a variety of scientific disciplines. The advent of machine learning and the increase of computing power has increased the interest in the analysis and statistical testing of multidimensional data. We extend the powerful Kolmogorov-Smirnov two sample test to a high dimensional form in a similar manner to Fasano (Fasano, 1987). We call our result the d-dimensional Kolmogorov-Smirnov test (ddKS) and provide three novel contributions therewith: we develop an analytical equation for the significance of a given ddKS score, we provide an algorithm for computation of ddKS on modern computing hardware that is of constant time complexity for small sample sizes and dimensions, and we provide two approximate calculations of ddKS: one that reduces the time complexity to linear at larger sample sizes, and another that reduces the time complexity to linear with increasing dimension. We perform power analysis of ddKS and its approximations on a corpus of datasets and compare to other common high dimensional two sample tests and distances: Hotelling's T^2 test and Kullback-Leibler divergence. Our ddKS test performs well for all datasets, dimensions, and sizes tested, whereas the other tests and distances fail to reject the null hypothesis on at least one dataset. We therefore conclude that ddKS is a powerful multidimensional two sample test for general use, and can be calculated in a fast and efficient manner using our parallel or approximate methods. Open source implementations of all methods described in this work are located at https://github.com/pnnl/ddks.

Daniel Claborne, Eric Slyman, Karl Pazdernik

We train an identity verification architecture and evaluate modifications to the part of the model that combines audio and visual representations, including in scenarios where one input is missing in either of two examples to be compared. We report results on the Voxceleb1-E test set that suggest averaging the output embeddings improves error rate in the full-modality setting and when a single modality is missing, and makes more complete use of the embedding space than systems which use shared layers and discuss possible reasons for this behavior.

Marjolein Oostrom, Alex Hagen, Nicole LaHaye et al.

Understanding the relationship between the evolution of microstructures of irradiated LiAlO2 pellets and tritium diffusion, retention and release could improve predictions of tritium-producing burnable absorber rod performance. Given expert-labeled segmented images of irradiated and unirradiated pellets, we trained Deep Convolutional Neural Networks to segment images into defect, grain, and boundary classes. Qualitative microstructural information was calculated from these segmented images to facilitate the comparison of unirradiated and irradiated pellets. We tested modifications to improve the sensitivity of the model, including incorporating meta-data into the model and utilizing uncertainty quantification. The predicted segmentation was similar to the expert-labeled segmentation for most methods of microstructural qualification, including pixel proportion, defect area, and defect density. Overall, the high performance metrics for the best models for both irradiated and unirradiated images shows that utilizing neural network models is a viable alternative to expert-labeled images.

Sai Munikoti, Ian Stewart, Sameera Horawalavithana et al.

Multimodal models are expected to be a critical component to future advances in artificial intelligence. This field is starting to grow rapidly with a surge of new design elements motivated by the success of foundation models in natural language processing (NLP) and vision. It is widely hoped that further extending the foundation models to multiple modalities (e.g., text, image, video, sensor, time series, graph, etc.) will ultimately lead to generalist multimodal models, i.e. one model across different data modalities and tasks. However, there is little research that systematically analyzes recent multimodal models (particularly the ones that work beyond text and vision) with respect to the underling architecture proposed. Therefore, this work provides a fresh perspective on generalist multimodal models (GMMs) via a novel architecture and training configuration specific taxonomy. This includes factors such as Unifiability, Modularity, and Adaptability that are pertinent and essential to the wide adoption and application of GMMs. The review further highlights key challenges and prospects for the field and guide the researchers into the new advancements.

Lee Burke, Karl Pazdernik, Daniel Fortin et al.

Natural language processing (NLP) tasks (text classification, named entity recognition, etc.) have seen revolutionary improvements over the last few years. This is due to language models such as BERT that achieve deep knowledge transfer by using a large pre-trained model, then fine-tuning the model on specific tasks. The BERT architecture has shown even better performance on domain-specific tasks when the model is pre-trained using domain-relevant texts. Inspired by these recent advancements, we have developed NukeLM, a nuclear-domain language model pre-trained on 1.5 million abstracts from the U.S. Department of Energy Office of Scientific and Technical Information (OSTI) database. This NukeLM model is then fine-tuned for the classification of research articles into either binary classes (related to the nuclear fuel cycle [NFC] or not) or multiple categories related to the subject of the article. We show that continued pre-training of a BERT-style architecture prior to fine-tuning yields greater performance on both article classification tasks. This information is critical for properly triaging manuscripts, a necessary task for better understanding citation networks that publish in the nuclear space, and for uncovering new areas of research in the nuclear (or nuclear-relevant) domains.