Research Papers

In the dynamic landscape of artificial intelligence, the exploration of hallucinations within vision-language (VL) models emerges as a critical frontier. This work delves into the intricacies of hallucinatory phenomena exhibited by widely used image captioners, unraveling interesting patterns. Specifically, we step upon previously introduced techniques of conceptual counterfactual explanations to address VL hallucinations. The deterministic and efficient nature of the employed conceptual counterfactuals backbone is able to suggest semantically minimal edits driven by hierarchical knowledge, so that the transition from a hallucinated caption to a non-hallucinated one is performed in a black-box manner. HalCECE, our proposed hallucination detection framework is highly interpretable, by providing semantically meaningful edits apart from standalone numbers, while the hierarchical decomposition of hallucinated concepts leads to a thorough hallucination analysis. Another novelty tied to the current work is the investigation of role hallucinations, being one of the first works to involve interconnections between visual concepts in hallucination detection. Overall, HalCECE recommends an explainable direction to the crucial field of VL hallucination detection, thus fostering trustworthy evaluation of current and future VL systems.

The widespread adoption of machine learning systems has raised critical concerns about fairness and bias, making mitigating harmful biases essential for AI development. In this paper, we investigate the relationship between debiasing and removing artifacts in neural networks for computer vision tasks. First, we introduce a set of novel XAI-based metrics that analyze saliency maps to assess shifts in a model's decision-making process. Then, we demonstrate that successful debiasing methods systematically redirect model focus away from protected attributes. Finally, we show that techniques originally developed for artifact removal can be effectively repurposed for improving fairness. These findings provide evidence for the existence of a bidirectional connection between ensuring fairness and removing artifacts corresponding to protected attributes.

In the context of explainable artificial intelligence (XAI), limited research has identified role-specific explanation needs. This study investigates the explanation needs of data scientists, who are responsible for training, testing, deploying, and maintaining machine learning (ML) models in AI systems. The research aims to determine specific explanation content of data scientists. A task analysis identified user goals and proactive user tasks. Using explanation questions, task-specific explanation needs and content were identified. From these individual explanations, we developed a mental model for explanations, which was validated and revised through a qualitative study (n=12). In a second quantitative study (n=12), we examined which explanation intents (reason, comparison, accuracy, prediction, trust) require which type of explanation content from the mental model. The findings are: F1: Explanation content for data scientists comes from the application domain, system domain, and AI domain. F2: Explanation content can be complex and should be organized sequentially and/or in hierarchies (novelty claim). F3: Explanation content includes context, inputs, evidence, attributes, ranked list, interim results, efficacy principle, and input/output relationships (novelty claim). F4: Explanation content should be organized as a causal story. F5: Standardized explanation questions ensure complete coverage of explanation needs (novelty claim). F6: Refining mental models for explanations increases significantly its quality (novelty claim).

Obtaining high-quality explanations of a model's output enables developers to identify and correct biases, align the system's behavior with human values, and ensure ethical compliance. Explainable Artificial Intelligence (XAI) practitioners rely on specific measures to gauge the quality of such explanations. These measures assess key attributes, such as how closely an explanation aligns with a model's decision process (faithfulness), how accurately it pinpoints the relevant input features (localization), and its consistency across different cases (robustness). Despite providing valuable information, these measures do not fully address a critical practitioner's concern: how does the quality of a given explanation compare to other potential explanations? Traditionally, the quality of an explanation has been assessed by comparing it to a randomly generated counterpart. This paper introduces an alternative: the Quality Gap Estimate (QGE). The QGE method offers a direct comparison to what can be viewed as the `inverse' explanation, one that conceptually represents the antithesis of the original explanation. Our extensive testing across multiple model architectures, datasets, and established quality metrics demonstrates that the QGE method is superior to the traditional approach. Furthermore, we show that QGE enhances the statistical reliability of these quality assessments. This advance represents a significant step toward a more insightful evaluation of explanations that enables a more effective inspection of a model's behavior.

This paper presents the development of an Artificial Intelligence (AI) based fighter jet agent within a customized Pygame simulation environment, designed to solve multi-objective tasks via deep reinforcement learning (DRL). The jet's primary objectives include efficiently navigating the environment, reaching a target, and selectively engaging or evading an enemy. A reward function balances these goals while optimized hyperparameters enhance learning efficiency. Results show more than 80\% task completion rate, demonstrating effective decision-making. To enhance transparency, the jet's action choices are analyzed by comparing the rewards of the actual chosen action (factual action) with those of alternate actions (counterfactual actions), providing insights into the decision-making rationale. This study illustrates DRL's potential for multi-objective problem-solving with explainable AI. Project page is available at: \href{https://github.com/swatikar95/Autonomous-Fighter-Jet-Navigation-and-Combat}{Project GitHub Link}.

Computational Safety for Generative AI: A Hypothesis Testing Perspective

Pin-Yu Chen

AI safety is a rapidly growing area of research that seeks to prevent the harm and misuse of frontier AI technology, particularly with respect to generative AI (GenAI) tools that are capable of creating realistic and high-quality content through text prompts. Examples of such tools include large language models (LLMs) and text-to-image (T2I) diffusion models. As the performance of various leading GenAI models approaches saturation due to similar training data sources and neural network architecture designs, the development of reliable safety guardrails has become a key differentiator for responsibility and sustainability. This paper presents a formalization of the concept of computational safety, which is a mathematical framework that enables the quantitative assessment, formulation, and study of safety challenges in GenAI through the lens of signal processing theory and methods. In particular, we explore two exemplary categories of computational safety challenges in GenAI that can be formulated as hypothesis testing problems. For the safety of model input, we show how sensitivity analysis and loss landscape analysis can be used to detect malicious prompts with jailbreak attempts. For the safety of model output, we elucidate how statistical signal processing can be used to detect AI-generated content. Finally, we discuss key open research challenges, opportunities, and the essential role of signal processing in computational AI safety.

As artificial intelligence (AI) systems become increasingly integrated into critical domains, ensuring their responsible design and continuous development is imperative. Effective AI quality management (QM) requires tools and methodologies that address the complexities of the AI lifecycle. In this paper, we propose an approach for AI lifecycle planning that bridges the gap between generic guidelines and use case-specific requirements (MQG4AI). Our work aims to contribute to the development of practical tools for implementing Responsible AI (RAI) by aligning lifecycle planning with technical, ethical and regulatory demands. Central to our approach is the introduction of a flexible and customizable Methodology based on Quality Gates, whose building blocks incorporate RAI knowledge through information linking along the AI lifecycle in a continuous manner, addressing AIs evolutionary character. For our present contribution, we put a particular emphasis on the Explanation stage during model development, and illustrate how to align a guideline to evaluate the quality of explanations with MQG4AI, contributing to overall Transparency.

Bias issues of neural networks garner significant attention along with its promising advancement. Among various bias issues, mitigating two predominant biases is crucial in advancing fair and trustworthy AI: (1) ensuring neural networks yields even performance across demographic groups, and (2) ensuring algorithmic decision-making does not rely on protected attributes. However, upon the investigation of \pc papers in the relevant literature, we find that there exists a persistent, extensive but under-explored confusion regarding these two types of biases. Furthermore, the confusion has already significantly hampered the clarity of the community and subsequent development of debiasing methodologies. Thus, in this work, we aim to restore clarity by providing two mathematical definitions for these two predominant biases and leveraging these definitions to unify a comprehensive list of papers. Next, we highlight the common phenomena and the possible reasons for the existing confusion. To alleviate the confusion, we provide extensive experiments on synthetic, census, and image datasets, to validate the distinct nature of these biases, distinguish their different real-world manifestations, and evaluate the effectiveness of a comprehensive list of bias assessment metrics in assessing the mitigation of these biases. Further, we compare these two types of biases from multiple dimensions including the underlying causes, debiasing methods, evaluation protocol, prevalent datasets, and future directions. Last, we provide several suggestions aiming to guide researchers engaged in bias-related work to avoid confusion and further enhance clarity in the community.

Most commonly used non-linear machine learning methods are closed-box models, uninterpretable to humans. The field of explainable artificial intelligence (XAI) aims to develop tools to examine the inner workings of these closed boxes. An often-used model-agnostic approach to XAI involves using simple models as local approximations to produce so-called local explanations; examples of this approach include LIME, SHAP, and SLISEMAP. This paper shows how a large set of local explanations can be reduced to a small "proxy set" of simple models, which can act as a generative global explanation. This reduction procedure, ExplainReduce, can be formulated as an optimisation problem and approximated efficiently using greedy heuristics. We show that, for many problems, as few as five explanations can faithfully emulate the closed-box model and that our reduction procedure is competitive with other model aggregation methods.

My research explores integrating deep learning and logic programming to set the basis for a new generation of AI systems. By combining neural networks with Inductive Logic Programming (ILP), the goal is to construct systems that make accurate predictions and generate comprehensible rules to validate these predictions. Deep learning models process and analyze complex data, while ILP techniques derive logical rules to prove the network's conclusions. Explainable AI methods, like eXplainable Answer Set Programming (XASP), elucidate the reasoning behind these rules and decisions. The focus is on applying ILP frameworks, specifically ILASP and FastLAS, to enhance explainability in various domains. My test cases span weather prediction, the legal field, and image recognition. In weather forecasting, the system will predict events and provides explanations using FastLAS, with plans to integrate recurrent neural networks in the future. In the legal domain, the research focuses on interpreting vague decisions and assisting legal professionals by encoding Italian legal articles and learning reasoning patterns from Court of Cassation decisions using ILASP. For biological laboratories, we will collaborate with a research group to automate spermatozoa morphology classification for Bull Breeding Soundness Evaluation using YOLO networks and ILP to explain classification outcomes. This hybrid approach aims to bridge the gap between the high performance of deep learning models and the transparency of symbolic reasoning, advancing AI by providing interpretable and trustworthy applications.

When deployed in the real world, AI agents will inevitably face challenges that exceed their individual capabilities. A critical component of AI safety is an agent's ability to recognize when it is likely to fail in a novel situation and to yield control to a more capable expert system. Leveraging such expert assistance can significantly improve safety and performance in such situations. Since expert assistance is costly, a central challenge is determining when to consult an expert. In this paper, we explore a novel variant of this problem, termed YRC-0, in which an agent must learn to collaborate with an expert in new environments in an unsupervised manner--that is, without interacting with the expert during training. This setting motivates the development of low-cost, robust approaches for training expert-leveraging agents. To support research in this area, we introduce YRC-Bench, an open-source benchmark that instantiates YRC-0 across diverse environments. YRC-Bench provides a standardized Gym-like API, simulated experts, an evaluation pipeline, and implementations of popular baselines. Toward tackling YRC-0, we propose a validation strategy and use a proposer-validator decomposition as a diagnostic framework to evaluate a range of learning methods, offering insights that can inform future research. Codebase: https://github.com/modanesh/YRC-Bench

The recent surge in advanced generative models, such as diffusion models and generative adversarial networks (GANs), has led to an alarming rise in AI-generated images across various domains on the web. While such technologies offer benefits such as democratizing artistic creation, they also pose challenges in misinformation, digital forgery, and authenticity verification. Additionally, the uncredited use of AI-generated images in media and marketing has sparked significant backlash from online communities. In response to this, we introduce DejAIvu, a Chrome Web extension that combines real-time AI-generated image detection with saliency-based explainability while users browse the web. Using an ONNX-optimized deep learning model, DejAIvu automatically analyzes images on websites such as Google Images, identifies AI-generated content using model inference, and overlays a saliency heatmap to highlight AI-related artifacts. Our approach integrates efficient in-browser inference, gradient-based saliency analysis, and a seamless user experience, ensuring that AI detection is both transparent and interpretable. We also evaluate DejAIvu across multiple pretrained architectures and benchmark datasets, demonstrating high accuracy and low latency, making it a practical and deployable tool for enhancing AI image accountability. The code for this system can be found at https://github.com/Noodulz/dejAIvu.