Generative Large Language Models (LLMs) are capable of being in-context learners. However, the underlying mechanism of in-context learning (ICL) is still a major research question, and experimental research results about how models exploit ICL are not always consistent. In this work, we propose a framework for evaluating in-context learning mechanisms, which we claim are a combination of retrieving internal knowledge and learning from in-context examples by focusing on regression tasks. First, we show that LLMs can solve real-world regression problems and then design experiments to measure the extent to which the LLM retrieves its internal knowledge versus learning from in-context examples. We argue that this process lies on a spectrum between these two extremes. We provide an in-depth analysis of the degrees to which these mechanisms are triggered depending on various factors, such as prior knowledge about the tasks and the type and richness of the information provided by the in-context examples. We employ three LLMs and utilize multiple datasets to corroborate the robustness of our findings. Our results shed light on how to engineer prompts to leverage meta-learning from in-context examples and foster knowledge retrieval depending on the problem being addressed.

The expansion of the Internet of Things (IoT) in the battlefield, Internet of Battlefield Things (IoBT), gives rise to new opportunities for enhancing situational awareness. To increase the potential of IoBT for situational awareness in critical decision making, the data from these devices must be processed into consumer-ready information objects and made available to consumers on demand. To address this challenge we propose a workflow that makes use of natural language processing (NLP) to query a database and return a response in natural language. Our solution utilizes Large Language Models (LLMs) sized for edge devices to perform NLP, as well as a graph database. These types of databases are well suited for the dynamic, connected networks pervasive in the IoBT. Our architecture employs LLMs for both mapping questions in natural language to Cypher database queries as well as to summarize the database output back to the user in natural language. We evaluated several medium-sized LLMs for both of these tasks on a database representing publicly available data from the US Army's Multi-purpose Sensing Area Multi-Purpose Sensing Area (MSA) at the Jornada Range in Las Cruces, NM. We observe that Llama 3.1 (8 billion parameters) outperforms the other models across all considered metrics. Most importantly, we note that, unlike current methods, our two step approach allows the relaxation of the Exact Match (EM) requirement of the produced Cypher queries with ground truth code and, in this way, it achieves a 19.4% increase in accuracy. Our workflow lays the groundwork for deploying LLMs on edge devices to enable natural language interactions with databases containing information objects for critical decision making.

This paper considers the challenges Large Language Models (LLMs) face when reasoning over text that includes information involving uncertainty explicitly quantified via probability values. This type of reasoning is relevant to a variety of contexts ranging from everyday conversations to medical decision-making. Despite improvements in the mathematical reasoning capabilities of LLMs, they still exhibit significant difficulties when it comes to probabilistic reasoning. To deal with this problem, we introduce the Bayesian Linguistic Inference Dataset (BLInD), a new dataset specifically designed to test the probabilistic reasoning capabilities of LLMs. We use BLInD to find out the limitations of LLMs for tasks involving probabilistic reasoning. In addition, we present several prompting strategies that map the problem to different formal representations, including Python code, probabilistic algorithms, and probabilistic logical programming. We conclude by providing an evaluation of our methods on BLInD and an adaptation of a causal reasoning question-answering dataset. Our empirical results highlight the effectiveness of our proposed strategies for multiple LLMs.

In this work, we evaluate the potential of Large Language Models (LLMs) in building Bayesian Networks (BNs) by approximating domain expert priors. LLMs have demonstrated potential as factual knowledge bases; however, their capability to generate probabilistic knowledge about real-world events remains understudied. We explore utilizing the probabilistic knowledge inherent in LLMs to derive probability estimates for statements regarding events and their relationships within a BN. Using LLMs in this context allows for the parameterization of BNs, enabling probabilistic modeling within specific domains. Our experiments on eighty publicly available Bayesian Networks, from healthcare to finance, demonstrate that querying LLMs about the conditional probabilities of events provides meaningful results when compared to baselines, including random and uniform distributions, as well as approaches based on next-token generation probabilities. We explore how these LLM-derived distributions can serve as expert priors to refine distributions extracted from data, especially when data is scarce. Overall, this work introduces a promising strategy for automatically constructing Bayesian Networks by combining probabilistic knowledge extracted from LLMs with real-world data. Additionally, we establish the first comprehensive baseline for assessing LLM performance in extracting probabilistic knowledge.

Effective operation and seamless cooperation of robotic systems are a fundamental component of next-generation technologies and applications. In contexts such as disaster response, swarm operations require coordinated behavior and mobility control to be handled in a distributed manner, with the quality of the agents' actions heavily relying on the communication between them and the underlying network. In this paper, we formulate the problem of dynamic network bridging in a novel Decentralized Partially Observable Markov Decision Process (Dec-POMDP), where a swarm of agents cooperates to form a link between two distant moving targets. Furthermore, we propose a Multi-Agent Reinforcement Learning (MARL) approach for the problem based on Graph Convolutional Reinforcement Learning (DGN) which naturally applies to the networked, distributed nature of the task. The proposed method is evaluated in a simulated environment and compared to a centralized heuristic baseline showing promising results. Moreover, a further step in the direction of sim-to-real transfer is presented, by additionally evaluating the proposed approach in a near Live Virtual Constructive (LVC) UAV framework.

We consider a scenario where a user must make a set of correlated decisions and we propose a computational cognitive model of the deliberation process. We assume the user compactly expresses her preferences via soft constraints and we study how a psychology-based model of human decision-making, namely Multi-Alternative Decision Field Theory (MDFT), can be applied in this context. We design and study sequential and synchronous procedures which combine local decision-making on each variable, with constraint propagation, as well as a one-shot approach. Our experimental results, which focus on tree-shaped Fuzzy Constraint Satisfaction Problems, suggest that decomposing the decision process along the preference structure allows to find solutions of high quality in terms of preferences, maintains MDFT’s ability to replicate behavioral effects and is more efficient in terms of computational cost.

Efficient information dissemination is crucial for supporting critical operations across domains like disaster response, autonomous vehicles, and sensor networks. This paper introduces a Multi-Agent Reinforcement Learning (MARL) approach as a significant step forward in achieving more decentralized, efficient, and collaborative information dissemination. We propose a Partially Observable Stochastic Game (POSG) formulation for information dissemination empowering each agent to decide on message forwarding independently, based on the observation of their one-hop neighborhood. This constitutes a significant paradigm shift from heuristics currently employed in real-world broadcast protocols. Our novel approach harnesses Graph Convolutional Reinforcement Learning and Graph Attention Networks (GATs) with dynamic attention to capture essential network features. We propose two approaches, L-DyAN and HL-DyAN, which differ in terms of the information exchanged among agents. Our experimental results show that our trained policies outperform existing methods, including the state-of-the-art heuristic, in terms of network coverage as well as communication overhead on dynamic networks of varying density and behavior.

In recent years, leveraging satellite imagery with deep learning (DL) architectures has become an effective approach for environmental monitoring tasks, including forest wildfire detection. Nevertheless, this integration requires substantial high-quality labeled data to train the DL models accurately. Leveraging the capabilities of multiple Python libraries, such as rasterio and GeoPandas, and Google Earth Engine’s Python API, this study introduces a streamlined methodology to efficiently gather, label, augment, process, and evaluate a large-scale bi-temporal high-resolution satellite imagery dataset for DL-driven forest wildfire detection. Known as the California Wildfire GeoImaging Dataset (CWGID), this dataset comprises over 100,000 labeled ’before’ and ’after’ wildfire image pairs, created from pre-existing satellite imagery. An analysis of the dataset using pre-trained and adapted Convolutional Neural Network (CNN) architectures, such as VGG16 and EfficientNet, achieved accuracies of respectively 76% and 93%. The pipeline outlined in this paper demonstrates how Python can be used to gather and process high-resolution satellite imagery datasets, leading to accurate wildfire detection and providing a tool for broader environmental monitoring.

In this paper, we evaluate the capability of transformer-based language models in making inferences over uncertain text that includes uncertain rules of reasoning. We cover both Pre-trained Language Models (PLMs) and generative Large Language Models (LLMs). Our evaluation results show that both generations of language models struggle with reasoning over uncertain text. We propose a novel end-to-end fine-tuning approach, Probabilistic Constraint Training (PCT), that utilizes probabilistic logical rules as constraints in the fine-tuning phase without relying on these rules in the inference stage. To assess the effectiveness of PCT, we utilize the related corpora and, additionally, create a new and more challenging benchmark that, unlike the previous ones, uses instance-specific rules. Our study demonstrates that PCT improves the transformer-based language model's intrinsic reasoning and makes their probabilistic logical reasoning process more explicit and explainable. Furthermore, PCT equips these models to effectively handle novel situations, including higher reasoning depth, new domains, and complex probabilistic structures.

Artificial intelligence (AI) still lacks human capabilities, like adaptability, generalizability, self-control, consistency, common sense, and causal reasoning. Humans achieve some of these capabilities by carefully combining their thinking "fast" and "slow". In this work we define an AI architecture that embeds these two modalities, and we study the role of a "meta-cognitive" component, with the role of coordinating and combining them, in achieving higher quality decisions.

Modeling human decision making plays a fundamental role in the design of intelligent systems capable of rich interactions and effective teamwork. In this paper we consider the task of choice prediction in settings with multiple alternatives. Cognitive models of decision making can successfully replicate and explain behavioral effects involving uncertainty and interactions among alternatives but are computationally intensive to train. ML approaches excel in terms of choice prediction accuracy, but fail to provide insights on the underlying preference reasoning. We study different degrees of integration of ML and cognitive models for this task. We show, via testing on behavioral data, that our hybrid approach, based on the integration of a neural network and the Multi-alternative Linear Ballistic Accumulator cognitive model, requires significantly less time to train, and allows to capture important cognitive parameters while maintaining similar accuracy to the pure ML approach.

Many real-life scenarios require humans to make difficult trade-offs: do we always follow all the traffic rules or do we violate the speed limit in an emergency? In general, how should we account for and balance the ethical values, safety recommendations, and societal norms, when we are trying to achieve a certain objective? To enable effective AI-human collaboration, we must equip AI agents with a model of how humans make such trade-offs in environments where there is not only a goal to be reached, but there are also ethical constraints to be considered and to possibly align with. These ethical constraints could be both deontological rules on actions that should not be performed, or also consequentialist policies that recommend avoiding reaching certain states of the world. Our purpose is to build AI agents that can mimic human behavior in these ethically constrained decision environments, with a long term research goal to use AI to help humans in making better moral judgments and actions. To this end, we propose a computational approach where competing objectives and ethical constraints are orchestrated through a method that leverages a cognitive model of human decision making, called multi-alternative decision field theory (MDFT). Using MDFT, we build an orchestrator, called MDFT-Orchestrator (MDFT-O), that is both general and flexible. We also show experimentally that MDFT-O both generates better decisions than using a heuristic that takes a weighted average of competing policies (WA-O), but also performs better in terms of mimicking human decisions as collected through Amazon Mechanical Turk (AMT). Our methodology is therefore able to faithfully model human decision in ethically constrained decision environments.

Many real-life scenarios require humans to make difficult trade-offs: do we always follow all the traffic rules or do we violate the speed limit in an emergency? These scenarios force us to evaluate the trade-off between collective rules and norms with our own personal objectives and desires. To create effective AI-human teams, we must equip AI agents with a model of how humans make these trade-offs in complex environments when there are implicit and explicit rules and constraints. Agent equipped with these models will be able to mirror human behavior and/or to draw human attention to situations where decision making could be improved. To this end, we propose a novel inverse reinforcement learning (IRL) method: Max Entropy Inverse Soft Constraint IRL (MESC-IRL), for learning implicit hard and soft constraints over states, actions, and state features from demonstrations in deterministic and non-deterministic environments modeled as Markov Decision Processes (MDPs). Our method enables agents implicitly learn human constraints and desires without the need for explicit modeling by the agent designer and to transfer these constraints between environments. Our novel method generalizes prior work which only considered deterministic hard constraints and achieves state of the art performance.

The stable marriage problem (SMP) is a mathematical abstraction of two-sided matching markets with many practical applications including matching resident doctors to hospitals and students to schools. Several preference models have been considered in the context of SMPs including orders with ties, incomplete orders, and orders with uncertainty, but none have yet captured behavioral aspects of human decision making, e.g., contextual effects of choice. We introduce Behavioral Stable Marriage Problems (BSMPs), bringing together the formalism of matching with cognitive models of decision making to account for multi-attribute, non-deterministic preferences and to study the impact of well known behavioral deviations from rationality on two core notions of SMPs: stability and fairness. We analyze the computational complexity of BSMPs and show that proposal-based approaches are affected by contextual effects. We then propose and evaluate novel ILP and local-search-based methods to efficiently find optimally stable and fair matchings for BSMPs.

In many real world situations, collective decisions are made using voting and, in scenarios such as committee or board elections, employing voting rules that return multiple winners. In multi-winner approval voting (AV), an agent submits a ballot consisting of approvals for as many candidates as they wish, and winners are chosen by tallying up the votes and choosing the top-k candidates receiving the most approvals. In many scenarios, an agent may manipulate the ballot they submit in order to achieve a better outcome by voting in a way that does not reflect their true preferences. In complex and uncertain situations, agents may use heuristics instead of incurring the additional effort required to compute the manipulation which most favors them. In this paper, we examine voting behavior in single-winner and multi-winner approval voting scenarios with varying degrees of uncertainty using behavioral data obtained from Mechanical Turk. We find that people generally manipulate their vote to obtain a better outcome, but often do not identify the optimal manipulation. There are a number of predictive models of agent behavior in the social choice and psychology literature that are based on cognitively plausible heuristic strategies. We show that the existing approaches do not adequately model our real-world data. We propose a novel model that takes into account the size of the winning set and human cognitive constraints; and demonstrate that this model is more effective at capturing real-world behaviors in multi-winner approval voting scenarios.

We introduce PCP-nets, a formalism to model qualitative conditional preferences with probabilistic uncertainty. PCP-nets generalise CP-nets by allowing for uncertainty over the preference orderings. We define and study both optimality and dominance queries in PCP-nets, and we propose a tractable approximation of dominance which we show to be very accurate in our experimental setting. Since PCP-nets can be seen as a way to model a collection of weighted CP-nets, we also explore the use of PCP-nets in a multi-agent context, where individual agents submit CP-nets which are then aggregated into a single PCP-net. We consider various ways to perform such aggregation and we compare them via two notions of scores, based on well known voting theory concepts. Experimental results allow us to identify the aggregation method that better represents the given set of CP-nets and the most efficient dominance procedure to be used in the multi-agent context.

Many real-life scenarios require humans to make difficult trade-offs: do we always follow all the traffic rules or do we violate the speed limit in an emergency? These scenarios force us to evaluate the trade-off between collective norms and our own personal objectives. To create effective AI-human teams, we must equip AI agents with a model of how humans make trade-offs in complex, constrained environments. These agents will be able to mirror human behavior or to draw human attention to situations where decision making could be improved. To this end, we propose a novel inverse reinforcement learning (IRL) method for learning implicit hard and soft constraints from demonstrations, enabling agents to quickly adapt to new settings. In addition, learning soft constraints over states, actions, and state features allows agents to transfer this knowledge to new domains that share similar aspects. We then use the constraint learning method to implement a novel system architecture that leverages a cognitive model of human decision making, multi-alternative decision field theory (MDFT), to orchestrate competing objectives. We evaluate the resulting agent on trajectory length, number of violated constraints, and total reward, demonstrating that our agent architecture is both general and achieves strong performance. Thus we are able to capture and replicate human-like trade-offs from demonstrations in environments when constraints are not explicit.

Artificial Intelligence has witnessed an exponential growth in the last decade and, thanks to its many successful and pervasive applications, it has now become a research field with profound societal impacts. The interest in AI has reached an all-time high from all sectors of our modern society, including industry, health, education and government. AI Magazine, founded in 1980, has documented the rise of AI from an elite and almost esoteric field to its current status of key player in modern society. Under the leadership of exceptional scientists with a global vision of the field, David Leake first and then Ashok Goel, it has provided a venue for vibrant discussion on technological transformations, research trends and fundamental breakthroughs.

Preference are central to decision making by both machines and humans. Representing, learning, and reasoning with preferences is an important area of study both within computer science and across the sciences. When working with preferences it is necessary to understand and compute the distance between sets of objects, e.g., the preferences of a user and a the descriptions of objects to be recommended. We present CPDIST, a novel neural network to address the problem of learning to measure the distance between structured preference representations. We use the popular CP-net formalism to represent preferences and then leverage deep neural networks to learn a recently proposed metric function that is computationally hard to compute directly. CPDIST is a novel metric learning approach based on the use of deep siamese networks which learn the Kendal Tau distance between partial orders that are induced by compact preference representations. We find that CPDIST is able to learn the distance function with high accuracy and outperform existing approximation algorithms on both the regression and classification task using less computation time. Performance remains good even when CPDIST is trained with only a small number of samples compared to the dimension of the solution space, indicating the network generalizes well.