United States

Artificial intelligence (AI) holds great promise in accelerating information processing in ways that can potentially help users make high-consequence decisions rapidly in operational environments. A key practical challenge is determining how to embed AI into command and control (C2) systems, especially those that are disconnected and have limited processing power. While large language models (LLMs) have demonstrated strong performance across various reasoning tasks, their size and computational requirements limit their suitability for deployment at the edge. The rapid growth of open-source small language models (SLMs) offers new opportunities for edge AI, but selecting the most appropriate model remains challenging, as standard benchmarks often fail to capture domain-specific tasks. This work reported here examines the viability of SLMs as support tools that enable users to make rapid decisions at the edge. As part of an Office of Naval Research-sponsored effort, we developed a novel evaluation framework combining insights from interviews with subject matter experts, workflow analyses, and guidance from naval doctrine to construct synthetic datasets and benchmarks representative of operational needs. 

We applied this framework to decision-making processes within the domain of operational law to evaluate the performance of SLMs in helping deployed commanders adhere to domestic and international regulations. While legal experts maintain the ultimate responsibility and authority in reviewing SLM outputs and making decisions, our results reveal benefits of SLM support for legal information retrieval and rules of engagement evaluation. 

These findings demonstrate that, with appropriate guardrails, SLMs provide a practical means to enhance decision support in resource-constrained C2 systems, improving both speed and reliability of critical actions. Finally, the proposed framework is generalizable to other mission domains and supports the development of AI-enabled C2 capabilities at the tactical edge. The poster will include illustrative workflow examples and SLM-assisted compliance checks to highlight practical operational applications.

NAML 2026

Evaluating the Efficacy of Small Language Models as Decision Support Tools at the Edge

ai benchmarks

human factors engineering

user-centered design

vector database

edge computing

### Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

Message from the General Chair: 
*It is my great pleasure and honor to welcome you to the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), held in beautiful Vienna, Austria, from July 27 to August 1, 2025. ACL2025continues our field’s tradition of excellence in scholarship, innovation, and inclusivity, and I am deeply grateful to the many volunteers who have worked tirelessly to bring this event to life.* 
[Read more](https://drive.google.com/file/d/1GI_hvOpjswAuYdUTromfeDiPpCcqidwg/view?usp=sharing)

To access this event page, you need to log in with the **email address you registered with**. Access credentials will be sent to your email from Underline - subject line "Welcome to ACL 2025". Please be sure to check your spam email folder if you do not see an email confirmation right away.

Please log in

To access this event page, you need to log in with the **email address you registered with**. 

ACL 2025

Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

poster

### Welcome to the 10th annual workshop on Naval Applications of Machine Learning!

NAML is the premier annual event showcasing current machine learning research relevant to Naval applications. We welcome participation from the defense community, industry, and academia. The NAML workshop is an opportunity for researchers from Navy research and development centers and our partner organizations to exchange technical ideas and to come together as a team. 
Naval Information Warfare Center Pacific is partnering with AFCEA San Diego chapter to host the workshop.
#### On-Demand Access

* This is an on-demand site and all the recorded content from the workshop is available here. 
* Access to the recordings is exclusively available to registered participants.
* Each registrant has received an email notification, with instructions on how to access the content.

[![](https://assets.underline.io/markdown_image/1/image/60cd12aa376c191d3467ea2ee60944ea.png)](https://naml2026.org/)

The goals of NAML are:

* To spread awareness of current machine learning research relevant to Naval applications
* To connect machine learning researchers with experts in military requirements
* To build and strengthen collaborations within the defense research community

#### *"What we do as a Navy is a human business. Don’t pass up an opportunity to come together as a team. Don’t pass up leadership opportunities, and don’t outsource them to a PowerPoint. If you have an opportunity to come together as a team, to increase belonging, do it."*
Admiral John M. Richardson, 31st Chief of Naval Operations

This is an on-demand site. All recorded content from NAML 2026 is available here and access to it is exclusively available to registered participants.

**Please log in using the email address you registered with.** If you have any issues logging in, please contact us at <naml2026@underline.io>.

Let’s celebrate the past decade by investing in the next. 🚀

Please log in to explore this event.

This is an on-demand site. All recorded content from NAML 2026 is available here and access to it is exclusively available to registered participants.

**Explore the latest advancements in machine learning tailored to Naval applications 🔎**

With on-demand access to NAML 2026, you can watch all lectures & browse poster presentations at your convenience, while also engaging directly with presenters via the Underline platform - ask questions, exchange ideas, and stay connected.

**If you still haven’t registered, you can do so now.**
👇
**Take advantage of a special one-time offer: full access for just $99, with complimentary registration for students, government, and military employees.**

**Get access here: https://naml2026-ondemand.paperform.co/**

Please check your spam folder, or contact our support team at **<naml2026@underline.io>** if you registered, but still haven't received an access email.

Let’s celebrate the past decade by investing in the next. 🚀

Please register!

The 10th Annual Workshop on Naval Applications of Machine Learning (NAML 2026) was held in San Diego, California, March 2-5, 2026. NAML is a forum for interdisciplinary experts from government, industry, and academia to discuss results of ongoing research on theory and applications of machine learning to address some of the most challenging problems faced by our naval forces. 



Sound propagation in the ocean depends on the parameterization of the seabed. The physical seabed properties are correlated and impact the data in different ways. To quantify the information content about the seabed properties in measured data, three tools for information geometry are used. These three methods are illustrated using bottom loss. Bottom loss represents the reflection coefficient of the seabed on a logarithmic scale and depends on both frequency and grazing angle. An analytical, multi-layered bottom loss model is used where the seabed layers are parameterized by their sound speed and density. Some combinations of layer properties produce similar bottom loss results, even if the specific values of their parameters are different. This ambiguity is investigated by applying information geometry to the model manifold: bottom loss as a function of frequency and grazing angle over a range of seabed properties. The resulting 2d model manifold is embedded in a data space with dimensions equal to the number of frequencies by the number of grazing angles. The edges of the model manifold correspond to bottom loss at the extreme values of one of the seabed parameters. To visualize this high dimensional data space, three information geometry techniques are used. For the first method, we use the Euclidean length of the edges of the manifold to indicate how much a parameter contributes to the overall model, which is referred to as “sloppy” or “stiff” parameters. The second method is to perform a Principal Component Analysis on the model manifold. Generally the first two principal components capture more than 90% of the model’s information and provide a simple way to evaluate the main sources of variation on the model manifold. The final method is to examine the Fisher Information Matrix (FIM). The FIM gives a quantitative measure of how much information a model carries about each parameter. To find the FIM, we perform automatic differentiation on the model for specific seabed properties. Using the FIM, we can find the conditional variance, Cramer-Rao bounds, and parameter correlations which can be examined to find how much information a parameter carries about the model. These information geometry techniques can be applied broadly to many types of models to quantify information content.

Information Geometry and Seafloor Bottom Loss

In this work, we present a multi-modal autonomous docking framework designed to operate robustly under unknown and changing dynamic conditions. The system supports three high-level operating modes: 1) human-in-the-loop supervision, 2) shared or assisted control, and 3) full autonomy, allowing seamless transitions across levels of operator involvement to support maritime operations 

Our approach integrates a deep-learning-based Koopman operator to obtain an accurate higher-dimensional linear representation of the vessel dynamics, enabling predictive control through an adaptive MPC formulation. Model mismatches and dynamic changes are handled using recursive least squares updates. Safety is enforced through a differentiable optimization–based Control Barrier Function (CBF) filter, which ensures real-time safe velocity commands across all modes of operation. 

Dock detection and localization are performed using multiple AprilTags, providing fast and reliable estimation of the docking structure and nearby obstacles. A simulation environment built in ROS and Gazebo enables systematic and scalable evaluation under a wide range of conditions. 

The proposed multi-modal control architecture is validated in simulation and through field experiments. Results indicate clear quantitative and qualitative improvements compared to baseline human drivers and alternative control strategies, demonstrating effective integration of differentiable CBFs with deep Koopman-based adaptive MPC for safe maritime docking.

Multi-Modal Safe Docking via Deep Koopman-Based Adaptive MPC and Differentiable CBFs

Biofilms form when ship hulls encounter naturally occurring microbial communities in marine waters. As these biofilms mature, their structure and microbiome evolve into a complex, self-sustaining microenvironment capable of supporting larger organisms such as barnacles and mussels. Environmental conditions (e.g., salinity, temperature, and water flow) strongly influence both the microscopic and macroscopic characteristics of developing biofilms. Because of this, biofilm composition can serve as a valuable indicator of geolocation. 

Raman microscopy provides a powerful means of examining biofilm microstructure by capturing vibrational signatures of molecular constituents. This technique yields non-destructive spectral fingerprints that are sensitive to spatial and developmental variations within biofilms, enabling precise identification. Traditional machine-learning approaches perform well on Raman data, allowing classification based on species composition, biofilm age, and responses to chemical or pharmacological stimuli. 

In this work, we present a method for determining environmental origin by distinguishing freshwater-derived biofilms from those grown in marine conditions. Although certain Raman spectral differences caused by salinity are known, the effectiveness of deep learning in detecting these subtle biochemical shifts remains largely unexplored. To investigate this, biofilms are being cultivated in both freshwater and seawater systems, with a target of 1,000 samples per condition. Each sample will be imaged, averaged to improve signal-to-noise ratio, and minimally preprocessed using baseline correction, smoothing, normalization, and cosmic-ray removal. The resulting spectra will be analyzed using a shallow 1D convolutional neural network to perform binary classification between freshwater and saltwater origins. Outcomes from this study will advance our broader goal of geolocating vessels based on biofilm signatures, with future plans to incorporate additional ecological influences on biofilm development. 



Classifying Freshwater and Marine Biofilms Using Raman Spectroscopy

This work covers the continued development of a concentration estimator for a contaminant plume generated by an unknown source in an underwater domain. The plant includes a stationary or mobile source, and the resulting plume is described using the 3D advection-diffusion equation. An autonomous mobile sensor moves through the plume and uses a process-state hybrid estimator to make real-time predictions of the plume from measurements of the concentration. The plant and estimator are discretized using a Non-Overlapping Domain Decomposition Explicit Finite Volume Method (NODDE-FVM) with constant time 4th order Runge-Kutta integration. Additionally, a novel method for measuring the gradient of a scalar field with a single mobile sensor is presented. This method is utilized by the mobile sensor to make estimations of the gradient of the plume without any information required from other sensors that may be present in the domain. This gradient evaluation method is based on a Taylor expansion of five or more scalar field measurements taken by the sensor along a 3D trajectory. The gradient results depend on both the conditioning of the resulting matrix equation and the error resulting from a Taylor series approximation of the field. Ongoing development of this work involves parameter estimation of the plume, focusing on source strength/release rate and localization in real time. This parameter estimation will follow similar work in implementing a neural-network framework to solve inverse problems for PDEs. In particular, we will utilize a physics-informed foundation to train a network on a general set of advective-diffusive scalar fields. This estimation scheme will be applied in tandem with the previously developed concentration estimator to complete a system that can estimate in real time the size, strength, and source location of a plume of contaminants in an underwater domain.

Real-Time Underwater Plume Estimation with a Mobile Sensor

Artificial intelligence models are increasingly integrated into naval systems for navigation, surveillance, decision support etc. These models can be targeted by adversarial attacks that can cause misclassifications. They also must perform under changing or degraded conditions such as lighting variation and changes in environment. In scenarios where AI is relied on for situational awareness and threat detection, the systems must continue to operate and provide credible outputs even in the presence of deliberate disruption.

This presentation describes a real-time defense and monitoring framework that keeps AI models detecting objects and targets under adversarial or degraded conditions while also flagging deliberate attacks. The system combines lightweight runtime analytics that track feature and confidence consistency across frames, anomaly detection that recognizes attack patterns in camera inputs, and analysis that separates adversarial interference from environmental noise. Our NAML 2025 presentation focused on fine-tuning models for improved robustness prior to deployment; this year’s work introduces an at-runtime defense that allows models to detect attacks in real-time.

Our team developed this work to meet the evaluation objectives of the Robust Artificial Intelligence Test Event (RAITE), a red-versus-blue exercise led by NSWC Crane that assesses AI robustness against cyber and physical threats. RAITE is scheduled for March 7-10, and internal evaluations aligned with its scenarios show that the system sustains object detection performance and identifies ongoing interference.

Our talk will share lessons learned in detecting and mitigating attacks on computer vision models, showing how to achieve real-time resilience without fine-tuning or major model changes. We will also discuss applications for shipboard surveillance, autonomous sensing, and expeditionary intelligence, surveillance, and reconnaissance (ISR) systems. All content is unclassified.

Layer-Wise Runtime Detection of Adversarial Attacks in Operational AI Systems

Automatic Target Recognition (ATR) tasks increasingly require multi-class identification of unmanned aerial vehicles (UAVs). However, existing computer vision training datasets contain only a few different types of airframes and are limited in geometric, lighting, and sensor diversity. Furthermore, most existing datasets provide only a single class label. As a result, when a new or custom UAV model must be identified — one that existing datasets do not contain — one must generate training datasets containing those models. Real-world data collection is typically prohibitive in terms of cost and effort. 
To address this, we present SYGARD (Synthetic Data Generator for Automatic Target Recognition, Identification, and Detection), a synthetic-data tool that enables fine-grained UAV classification and identification by ingesting a 3D model of any UAV and generating large volumes of labeled, photorealistic, geospatially grounded imagery under diverse environmental and sensor conditions. SYGARD integrates Unreal Engine 5 rendering, AirSim-based dynamics, Cesium geospatial environments, and automated ground truth generation within a modular workflow for mission design, camera configuration, multi-agent simulation, and real-to-sim scenario reconstruction to produce synthetic data that reflects realistic deployment conditions.
Using these capabilities, we validate SYGARD on an end-to-end identification task involving a custom fixed-wing UAV platform and a quadcopter: synthetic data generated from their 3D models is used to train a YOLO-based classifier that reliably distinguishes the custom airframe in real flight footage without any real training images. In addition, we show that integrating SYGARD data with real imagery through hybrid training improves cross-dataset UAV detection. SYGARD offers a practical and scalable foundation for future-ready ATR systems.

SYGARD: Synthetic Data Generator for Automatic Target Recognition, Identification, and Detection

Large Reasoning Models (LRMs) demonstrate strong performance on complex reasoning tasks, yet they often suffer from overthinking—producing excessively long chain-of-thought (CoT) traces that increase inference cost and may degrade accuracy. Our analysis reveals a clear anti-correlation between reasoning length and accuracy, where across multiple stochastic decodes, the short reasoning paths consistently achieve the highest correctness, while longer ones accumulate errors and repetitions. These short optimal reasoning paths can be found ideally through full enumeration of the reasoning space. However, the tree-structured reasoning space grows exponentially with sequence length, rendering exhaustive exploration infeasible. To address this, we propose DTS, a model-agnostic decoding framework that sketches the reasoning space by selectively branching at high-entropy tokens and applies early stopping to select the shortest completed reasoning path. This approach approximates the optimal solution that enhances both efficiency and accuracy, without requiring additional training or supervision. Experiments on AIME2024 and AIME2025 datasets with DeepSeek-R1-Distill-Qwen-7B and 1.5B show that DTS improves accuracy by up to 8%, reduces average reasoning length by 23%, and decreases repetition frequency by 12%, demonstrating DTS’s ability for scalable and efficient LRM reasoning.

DTS: Enhancing Large Reasoning Models via Decoding Tree Sketching

Subject: 

Creation of a Multidimensional Multimodal Anomaly Detection with Interpretability (MADI) model using in-flight sensor data to augment fault prediction and fleet health monitoring predictive maintenance modeling efforts for the F-15E aircraft. 



Abstract:  

This work presents an end-to-end MADI prototype developed to enhance anomaly detection capabilities using F-15E Flight Data. Flight data was ingested and transformed within a specialized preprocessing pipeline designed to stabilize raw flight data, align disparate recordings, and support a repeatable anomaly detection framework. To address the absence of labeled fault data in operational settings, synthetic anomalies were generated using a uniform negative-sampling procedure, enabling supervised model training without dependence on historical failure events. 

A custom deep neural network was trained and tuned using a reproducible MLOps framework with hyperparameter optimization and automated experiment tracking. The model produces per-timestamp anomaly probabilities and applies integrated-gradients attribution to identify which sensor channels most strongly contribute to each detected anomaly. This approach treats the ensemble state of all sensors simultaneously and enables detection of complex feature-cross anomalies missed by univariate and rule-based algorithms. These outputs are aggregated into sortie-level metrics that summarize anomaly prevalence, severity, and subsystem-level contribution patterns. Results are integrated into a Historical Maintenance Timeline to support alignment with flight activity, maintenance actions, and readiness events. This prototype demonstrates a scalable, explainable, and operationally relevant anomaly detection capability for fighter aircraft sensor data. The framework supports DoD objectives in condition-based maintenance, fleet health monitoring, and data-driven sustainment planning, and establishes a foundation for future enhancements including anomaly-sequence contraction, event correlation, subsystem mapping, and operational-mode-aware anomaly models.

MADI-Based Anomaly Detection to Enhance Predictive Maintenance for the F-15E

The Results-Aligned Performance Improvement & Development (RAPID) training system transforms training content creation by seamlessly blending artificial intelligence with human expertise. This innovative solution enables Naval commands to quickly create and update training materials when operational procedures and Naval policies change, ensuring fleet readiness. Our presentation will demonstrate a clickable prototype showcasing how RAPID can be deployed for critical naval training needs such as shipboard damage control procedures, tactical operations, and maintenance protocols where accuracy and compliance with Navy standards are paramount. Attendees will observe how instructional systems designers (ISDs) configure parameters, upload Navy-specific documentation, and generate draft content that both ISDs and subject matter experts (SMEs) then review and refine. RAPID stands apart from competitors in five key ways: 1) Unlike commercial AI solutions requiring cloud connectivity, RAPID can be rapidly deployed within secure networks, with no external dependencies, making it ideal for fleet-wide deployment. 2) While other systems focus on automating content creation, RAPID’s human-in-the-loop workflow preserves the ISD’s and SME’s expertise while ensuring critical human oversight. 3) The system’s parameter pack assistant allows users to customize and standardize the output logic and format of AI-generated content in compliance with Navy training standards, learning taxonomies, and domain-specific constraints across different learning contexts. 4) RAPID employs a specialized retrieval-augmented generation (RAG) architecture that can index Navy-specific documentation, ensuring all generated training content maintains fidelity to authorized naval sources and terminology. 5) The system's AI validation engine quantitatively measures content against customer-defined standards using proprietary algorithms for coverage (ensuring all objectives are addressed), groundedness (verifying factual accuracy), congruency (confirming alignment between objectives, content, and assessments), and measurability (ensuring outcomes can be properly evaluated).

RAPID Training System

Humanitarian Assistance and Disaster Relief (HADR) operations require rapid damage assessment to effectively allocate resources and save lives, yet traditional satellite imagery analysis can take days to complete. This delay creates a critical gap between disaster occurrence and informed response, costing lives and resources. Our team at CMU's Software Engineering Institute (SEI) presents a modernized, cloud-based platform that enables domain experts and mission users to deploy custom computer vision models for damage assessment within minutes rather than days, significantly accelerating response times across multiple Areas of Responsibility (AORs).
Our approach addresses a fundamental challenge in operational disaster response: the need to quickly adapt pre-trained models to new, out-of-distribution imagery from emerging incidents with diverse geographic, structural, and disaster characteristics. CMU SEI previously co-hosted the xView2 Prize Challenge and deployed machine learning solutions for EUCOM (Turkey & Ukraine, 2022-2024) and CENTCOM (Syria 2023). However, we observed critical limitations regarding fragmented workflows for model training and inferencing. Our platform leverages an image segmentation base model pre-trained on satellite imagery to predict damage classes across disaster types. When users encounter new scenarios, they upload imagery and rapidly train lightweight model adapters using LoRA (Low-Rank Adaptation), a Parameter-Efficient Fine-Tuning technique requiring limited computational resources and training data.
The adapter-based architecture provides several operational advantages: trained adapters can be reused across similar scenarios, fine-tuned for evolving situations, and shared across organizations. Domain experts can rapidly customize models for their specific AORs without deep machine learning expertise, democratizing access to advanced AI capabilities.
Our platform's cloud-agnostic architecture enables deployment on both commercial and government cloud environments and on-premises infrastructure, operating seamlessly from tactical edge environments to strategic operations centers, enabling cross-echelon collaboration. This infrastructure-agnostic design ensures damage assessment capabilities remain available regardless of network constraints or deployment environment, making it a force multiplier for operations where interoperability and rapid information sharing are paramount.
Preliminary results demonstrate that providing limited (2-16) labels to post-disaster satellite imagery yields substantial performance improvements over the base model, with low computational costs measured in minutes of GPU time. This efficiency is critical for first-wave response scenarios where every minute impacts mission outcomes. By enabling iterative refinement as new imagery arrives, our platform empowers Battle Damage Assessment to make data-driven resource allocation decisions at the speed required by modern HADR operations.

Premium content

Next from NAML 2026

Information Geometry and Seafloor Bottom Loss

Similar lecture

Demystifying Small Language Models for Edge Deployment