arXiv subjects

Sameera Horawalavithana

Publications and source records attributed to Sameera Horawalavithana.

At least 19 recordsLinked to original sources

arXiv2026

SCICONVBENCH: Benchmarking LLMs on Multi-Turn Clarification for Task Formulation in Computational Science

Large Language Models (LLMs) are increasingly deployed as scientific AI as- sistants, and a growing body of benchmarks evaluates their capabilities across knowledge retrieval, reasoning, code generation, and tool use. These evaluations, however, typically assume the scientific problem is already well-posed, whereas practical scientific assistance often begins with an ill-posed user request that must be refined through dialogue before any computation, analysis, or experiment can be carried out reliably. We introduce SCICONVBENCH, a benchmark for multi- turn clarification in scientific task formulation across four computational science problem domains: fluid mechanics, solid mechanics, materials science, and par- tial differential equations (PDEs). SCICONVBENCH targets two complementary capabilities: eliciting missing information (disambiguation) and detecting and correcting erroneous requests containing internally contradictory information (in- consistency resolution). Our benchmark pairs a structured task ontology with a rubric-based evaluation framework, enabling systematic measurement of LLM per- formance across three dimensions: clarification behavior, conversational grounding, and final-specification fidelity. Current frontier models perform relatively well on inconsistency resolution, but even the best model resolves only 52.7% of the disambiguation cases in fluid mechanics. We further find that frontier LLMs fre- quently make silent assumptions and perform implicit specification repairs that are not grounded in the conversation with users. SCICONVBENCH establishes a foundation for evaluating the upstream conversational reasoning that a reliable computational science assistant requires. The code and data can be found at https://github.com/csml-rpi/SciConvBench.

cs.AI↗

arXiv2026

Position: AI Agents in Scientific Teams Should Be Studied as Human-Agent Systems

Large language model-based agents are increasingly deployed as collaborators in scientific discovery yet most current work focuses on the autonomous capabilities of "AI Scientists". We argue that this overlooks the social aspects of scientific teamwork, and that studying AI Scientists as human-agent systems (HAS)--where the unit of analysis is the human-agent pair--is both underexplored and undervalued. We establish these points through literature and empirical analysis, and highlight recent incidences and studies which show that deploying agents in science without accounting for human-agent dynamics introduces near-term risks, including reduced diversity of scientific inquiry. Through analysis of real-world case studies, we show that scientists and agents can augment each other's capabilities. We call for new research that adopts the HAS lens to develop mathematical frameworks for understanding and fostering human-AI synergy in scientific discovery.

cs.AI↗

arXiv2026

MULTISEISMO: A Multimodal Seismic Dataset and Model for Cross-Modal Seismic Understanding

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.

cs.LG↗

arXiv2026

Evaluating Memory Condensation Strategies for Coding Agents in Data-Driven Scientific Discovery

Coding agents accumulate extensive context during long-running tasks, yet fixed context windows force practitioners to choose between truncation and task failure. While numerous memory condensation strategies have been proposed, from simple sliding windows to LLM-generated summaries, no systematic comparison exists to guide strategy selection, especially in scientific discovery tasks. We evaluate eight memory condensation strategies using GPT-4o on sixty DiscoveryBench tasks spanning six scientific domains (480 total evaluations). We find that no condenser significantly alters hypothesis quality, while LLM-based condensers increase token costs by 24-94 percent, and masking tool-call outputs achieves an 8.6 percent net savings. We also observe that the optimal condenser for data-driven scientific discovery varies by scientific domain and task length.

cs.LG↗

arXiv2026

Reward Design for Physical Reasoning in Vision-Language Models

Physical reasoning over visual inputs demands tight integration of visual perception, domain knowledge, and multi-step symbolic inference. Yet even state-of-the-art Vision Language Models (VLMs) fall far short of human performance on physics benchmarks. While post-training algorithms such as Supervised Fine-Tuning (SFT) and Group Relative Policy Optimization (GRPO) have demonstrated strong reasoning gains in language models, how reward design shapes VLM physical reasoning behavior remains poorly understood. We present a systematic reward ablation study for GRPO-based VLM training on physical reasoning. We compare four reward signals of increasing semantic richness: format compliance, answer accuracy, a composite rubric reward (answer correctness, physics principle identification, and unit consistency), and a novel internal reward derived from model attention weights over input image regions. We evaluate on PhyX, a 3,000-problem benchmark spanning six physics domains and six reasoning types across multiple-choice and open-ended formats, using IBM Granite Vision 3.3 (2B). Across both formats, GRPO with accuracy-based rewards outperforms SFT on most domains, though gains vary substantially by reward type and domain. Reward design does not uniformly improve performance. Instead, it induces domain-specific reasoning behaviors. Accuracy-based rewards provide the strongest overall gains. Rubric rewards improve structured reasoning quality without consistent accuracy improvements. Attention-based rewards enhance spatial reasoning while degrading performance in symbolic domains. Our internal attention-weight reward requires no spatial annotations and improves spatial relation accuracy from 0.27 to 0.50, suggesting that supervising where the model attends during generation is a promising direction for visually grounded physical reasoning.

cs.AI↗

arXiv2026

Back to the Barn with LLAMAs: Evolving Pretrained LLM Backbones in Finetuning Vision Language Models

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.

cs.AI↗

arXiv2026

SCITUNE: Aligning Large Language Models with Human-Curated Scientific Multimodal Instructions

Instruction finetuning is a popular paradigm to align large language models (LLM) with human intent. Despite its popularity, this idea is less explored in improving LLMs to align existing foundation models with scientific disciplines, concepts and goals. In this work, we present \textit{SciTune} as a tuning framework to improve the ability of LLMs to follow multimodal instructions generated from scientific publications. To test our methodology, we train a large multimodal model LLaMA-SciTune that connects a vision encoder and LLM for science-focused visual and language understanding. LLaMA-SciTune significantly outperforms the state-of-the-art models in the generated figure types and captions in SciCap and VisText benchmarks. In comparison to the models that are finetuned with synthetic data only, LLaMA-SciTune surpasses human performance on average and in many sub-categories on the ScienceQA benchmark. Our results demonstrate that human-generated scientific multimodal instructions remain highly valuable in tuning LLMs to perform well on science tasks, despite their lower volume and relative scarcity compared to synthetic data. We publicly release the SciTune codebase https://github.com/pnnl/scitune.

cs.CV↗

arXiv2025

Surprisingly Fragile: Assessing and Addressing Prompt Instability in Multimodal Foundation Models

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.

cs.CL↗

arXiv2025

Audit, Alignment, and Optimization of LM-Powered Subroutines with Application to Public Comment Processing

The advent of language models (LMs) has the potential to dramatically accelerate tasks that may be cast to text-processing; however, real-world adoption is hindered by concerns regarding safety, explainability, and bias. How can we responsibly leverage LMs in a transparent, auditable manner -- minimizing risk and allowing human experts to focus on informed decision-making rather than data-processing or prompt engineering? In this work, we propose a framework for declaring statically typed, LM-powered subroutines (i.e., callable, function-like procedures) for use within conventional asynchronous code -- such that sparse feedback from human experts is used to improve the performance of each subroutine online (i.e., during use). In our implementation, all LM-produced artifacts (i.e., prompts, inputs, outputs, and data-dependencies) are recorded and exposed to audit on demand. We package this framework as a library to support its adoption and continued development. While this framework may be applicable across several real-world decision workflows (e.g., in healthcare and legal fields), we evaluate it in the context of public comment processing as mandated by the 1969 National Environmental Protection Act (NEPA): Specifically, we use this framework to develop "CommentNEPA," an application that compiles, organizes, and summarizes a corpus of public commentary submitted in response to a project requiring environmental review. We quantitatively evaluate the application by comparing its outputs (when operating without human feedback) to historical ``ground-truth'' data as labelled by human annotators during the preparation of official environmental impact statements.

cs.CL↗

arXiv2025

Evaluating the Robustness of Dense Retrievers in Interdisciplinary Domains

Evaluation benchmark characteristics may distort the true benefits of domain adaptation in retrieval models. This creates misleading assessments that influence deployment decisions in specialized domains. We show that two benchmarks with drastically different features such as topic diversity, boundary overlap, and semantic complexity can influence the perceived benefits of fine-tuning. Using environmental regulatory document retrieval as a case study, we fine-tune ColBERTv2 model on Environmental Impact Statements (EIS) from federal agencies. We evaluate these models across two benchmarks with different semantic structures. Our findings reveal that identical domain adaptation approaches show very different perceived benefits depending on evaluation methodology. On one benchmark, with clearly separated topic boundaries, domain adaptation shows small improvements (maximum 0.61% NDCG gain). However, on the other benchmark with overlapping semantic structures, the same models demonstrate large improvements (up to 2.22% NDCG gain), a 3.6-fold difference in the performance benefit. We compare these benchmarks through topic diversity metrics, finding that the higher-performing benchmark shows 11% higher average cosine distances between contexts and 23% lower silhouette scores, directly contributing to the observed performance difference. These results demonstrate that benchmark selection strongly determines assessments of retrieval system effectiveness in specialized domains. Evaluation frameworks with well-separated topics regularly underestimate domain adaptation benefits, while those with overlapping semantic boundaries reveal improvements that better reflect real-world regulatory document complexity. Our findings have important implications for developing and deploying AI systems for interdisciplinary domains that integrate multiple topics.

cs.IR↗

arXiv2025

Benchmarking LLMs for Environmental Review and Permitting

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.

cs.CL↗

arXiv2025

WeQA: A Benchmark for Retrieval Augmented Generation in Wind Energy Domain

Wind energy project assessments present significant challenges for decision-makers, who must navigate and synthesize hundreds of pages of environmental and scientific documentation. These documents often span different regions and project scales, covering multiple domains of expertise. This process traditionally demands immense time and specialized knowledge from decision-makers. The advent of Large Language Models (LLM) and Retrieval Augmented Generation (RAG) approaches offer a transformative solution, enabling rapid, accurate cross-document information retrieval and synthesis. As the landscape of Natural Language Processing (NLP) and text generation continues to evolve, benchmarking becomes essential to evaluate and compare the performance of different RAG-based LLMs. In this paper, we present a comprehensive framework to generate a domain relevant RAG benchmark. Our framework is based on automatic question-answer generation with Human (domain experts)-AI (LLM) teaming. As a case study, we demonstrate the framework by introducing WeQA, a first-of-its-kind benchmark on the wind energy domain which comprises of multiple scientific documents/reports related to environmental aspects of wind energy projects. Our framework systematically evaluates RAG performance using diverse metrics and multiple question types with varying complexity level, providing a foundation for rigorous assessment of RAG-based systems in complex scientific domains and enabling researchers to identify areas for improvement in domain-specific applications.

cs.CL↗

arXiv2024

Generalist Multimodal AI: A Review of Architectures, Challenges and Opportunities

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.

cs.CL↗

arXiv2023

ATLANTIC: Structure-Aware Retrieval-Augmented Language Model for Interdisciplinary Science

Large language models record impressive performance on many natural language processing tasks. However, their knowledge capacity is limited to the pretraining corpus. Retrieval augmentation offers an effective solution by retrieving context from external knowledge sources to complement the language model. However, existing retrieval augmentation techniques ignore the structural relationships between these documents. Furthermore, retrieval models are not explored much in scientific tasks, especially in regard to the faithfulness of retrieved documents. In this paper, we propose a novel structure-aware retrieval augmented language model that accommodates document structure during retrieval augmentation. We create a heterogeneous document graph capturing multiple types of relationships (e.g., citation, co-authorship, etc.) that connect documents from more than 15 scientific disciplines (e.g., Physics, Medicine, Chemistry, etc.). We train a graph neural network on the curated document graph to act as a structural encoder for the corresponding passages retrieved during the model pretraining. Particularly, along with text embeddings of the retrieved passages, we obtain structural embeddings of the documents (passages) and fuse them together before feeding them to the language model. We evaluate our model extensively on various scientific benchmarks that include science question-answering and scientific document classification tasks. Experimental results demonstrate that structure-aware retrieval improves retrieving more coherent, faithful and contextually relevant passages, while showing a comparable performance in the overall accuracy.

cs.CL↗

arXiv2023

Empirical evaluation of Uncertainty Quantification in Retrieval-Augmented Language Models for Science

Large language models (LLMs) have shown remarkable achievements in natural language processing tasks, producing high-quality outputs. However, LLMs still exhibit limitations, including the generation of factually incorrect information. In safety-critical applications, it is important to assess the confidence of LLM-generated content to make informed decisions. Retrieval Augmented Language Models (RALMs) is relatively a new area of research in NLP. RALMs offer potential benefits for scientific NLP tasks, as retrieved documents, can serve as evidence to support model-generated content. This inclusion of evidence enhances trustworthiness, as users can verify and explore the retrieved documents to validate model outputs. Quantifying uncertainty in RALM generations further improves trustworthiness, with retrieved text and confidence scores contributing to a comprehensive and reliable model for scientific applications. However, there is limited to no research on UQ for RALMs, particularly in scientific contexts. This study aims to address this gap by conducting a comprehensive evaluation of UQ in RALMs, focusing on scientific tasks. This research investigates how uncertainty scores vary when scientific knowledge is incorporated as pretraining and retrieval data and explores the relationship between uncertainty scores and the accuracy of model-generated outputs. We observe that an existing RALM finetuned with scientific knowledge as the retrieval data tends to be more confident in generating predictions compared to the model pretrained only with scientific knowledge. We also found that RALMs are overconfident in their predictions, making inaccurate predictions more confidently than accurate ones. Scientific knowledge provided either as pretraining or retrieval corpus does not help alleviate this issue. We released our code, data and dashboards at https://github.com/pnnl/EXPERT2.

cs.CL↗

arXiv2023

Evaluating the Effectiveness of Retrieval-Augmented Large Language Models in Scientific Document Reasoning

Despite the dramatic progress in Large Language Model (LLM) development, LLMs often provide seemingly plausible but not factual information, often referred to as hallucinations. Retrieval-augmented LLMs provide a non-parametric approach to solve these issues by retrieving relevant information from external data sources and augment the training process. These models help to trace evidence from an externally provided knowledge base allowing the model predictions to be better interpreted and verified. In this work, we critically evaluate these models in their ability to perform in scientific document reasoning tasks. To this end, we tuned multiple such model variants with science-focused instructions and evaluated them on a scientific document reasoning benchmark for the usefulness of the retrieved document passages. Our findings suggest that models justify predictions in science tasks with fabricated evidence and leveraging scientific corpus as pretraining data does not alleviate the risk of evidence fabrication.

cs.CL↗

arXiv2023

NuclearQA: A Human-Made Benchmark for Language Models for the Nuclear Domain

As LLMs have become increasingly popular, they have been used in almost every field. But as the application for LLMs expands from generic fields to narrow, focused science domains, there exists an ever-increasing gap in ways to evaluate their efficacy in those fields. For the benchmarks that do exist, a lot of them focus on questions that don't require proper understanding of the subject in question. In this paper, we present NuclearQA, a human-made benchmark of 100 questions to evaluate language models in the nuclear domain, consisting of a varying collection of questions that have been specifically designed by experts to test the abilities of language models. We detail our approach and show how the mix of several types of questions makes our benchmark uniquely capable of evaluating models in the nuclear domain. We also present our own evaluation metric for assessing LLM's performances due to the limitations of existing ones. Our experiments on state-of-the-art models suggest that even the best LLMs perform less than satisfactorily on our benchmark, demonstrating the scientific knowledge gap of existing LLMs.

cs.CL↗

arXiv2023

Anticipating Technical Expertise and Capability Evolution in Research Communities using Dynamic Graph Transformers

The ability to anticipate technical expertise and capability evolution trends globally is essential for national and global security, especially in safety-critical domains like nuclear nonproliferation (NN) and rapidly emerging fields like artificial intelligence (AI). In this work, we extend traditional statistical relational learning approaches (e.g., link prediction in collaboration networks) and formulate a problem of anticipating technical expertise and capability evolution using dynamic heterogeneous graph representations. We develop novel capabilities to forecast collaboration patterns, authorship behavior, and technical capability evolution at different granularities (e.g., scientist and institution levels) in two distinct research fields. We implement a dynamic graph transformer (DGT) neural architecture, which pushes the state-of-the-art graph neural network models by (a) forecasting heterogeneous (rather than homogeneous) nodes and edges, and (b) relying on both discrete -- and continuous -- time inputs. We demonstrate that our DGT models predict collaboration, partnership, and expertise patterns with 0.26, 0.73, and 0.53 mean reciprocal rank values for AI and 0.48, 0.93, and 0.22 for NN domains. DGT model performance exceeds the best-performing static graph baseline models by 30-80% across AI and NN domains. Our findings demonstrate that DGT models boost inductive task performance, when previously unseen nodes appear in the test data, for the domains with emerging collaboration patterns (e.g., AI). Specifically, models accurately predict which established scientists will collaborate with early career scientists and vice-versa in the AI domain.

cs.LG↗