We study the efficacy of fine-tuning Large Language Models (LLMs) for the specific task of report (government archives, news, intelligence reports) summarization. While this topic is being very actively researched - our specific application set-up faces two challenges: (i) ground-truth summaries maybe unavailable (e.g., for government archives), and (ii) availability of limited compute power - the sensitive nature of the application requires that computation is performed on-premise and for most of our experiments we use one or two A100 GPU cards. Under this set-up we conduct experiments to answer the following questions. First, given that fine-tuning the LLMs can be resource intensive, is it feasible to fine-tune them for improved report summarization capabilities on-premise? Second, what are the metrics we could leverage to assess the quality of these summaries? We conduct experiments on two different fine-tuning approaches in parallel and our findings reveal interesting trends regarding the utility of fine-tuning LLMs. Specifically, we find that in many cases, fine-tuning helps improve summary quality and in other cases it helps by reducing the number of invalid or garbage summaries.

The lack of a common platform and benchmark datasets for evaluating face obfuscation methods has been a challenge, with every method being tested using arbitrary experiments, datasets, and metrics. While prior work has demonstrated that face recognition systems exhibit bias against some demographic groups, there exists a substantial gap in our understanding regarding the fairness of face obfuscation methods. Providing fair face obfuscation methods can ensure equitable protection across diverse demographic groups, especially since they can be used to preserve the privacy of vulnerable populations. To address these gaps, this paper introduces a comprehensive framework, named FairDeFace, designed to assess the adversarial robustness and fairness of face obfuscation methods. The framework introduces a set of modules encompassing data benchmarks, face detection and recognition algorithms, adversarial models, utility detection models, and fairness metrics. FairDeFace serves as a versatile platform where any face obfuscation method can be integrated, allowing for rigorous testing and comparison with other state-of-the-art methods. In its current implementation, FairDeFace incorporates 6 attacks, and several privacy, utility and fairness metrics. Using FairDeFace, and by conducting more than 500 experiments, we evaluated and compared the adversarial robustness of seven face obfuscation methods. This extensive analysis led to many interesting findings both in terms of the degree of robustness of existing methods and their biases against some gender or racial groups. FairDeFace also uses visualization of focused areas for both obfuscation and verification attacks to show not only which areas are mostly changed in the obfuscation process for some demographics, but also why they failed through focus area comparison of obfuscation and verification.

In the age of powerful AI-generated text, automatic detectors have emerged to identify machine-written content. This poses a threat to author privacy and freedom, as text authored with AI assistance may be unfairly flagged. We propose AuthorMist, a novel reinforcement learning-based system to transform AI-generated text into human-like writing. AuthorMist leverages a 3-billion-parameter language model as a backbone, fine-tuned with Group Relative Policy Optimization (GPRO) to paraphrase text in a way that evades AI detectors. Our framework establishes a generic approach where external detector APIs (GPTZero, WinstonAI, Originality.ai, etc.) serve as reward functions within the reinforcement learning loop, enabling the model to systematically learn outputs that these detectors are less likely to classify as AI-generated. This API-as-reward methodology can be applied broadly to optimize text against any detector with an accessible interface. Experiments on multiple datasets and detectors demonstrate that AuthorMist effectively reduces the detectability of AI-generated text while preserving the original meaning. Our evaluation shows attack success rates ranging from 78.6% to 96.2% against individual detectors, significantly outperforming baseline paraphrasing methods. AuthorMist maintains high semantic similarity (above 0.94) with the original text while successfully evading detection. These results highlight limitations in current AI text detection technologies and raise questions about the sustainability of the detection-evasion arms race.

Participants: This study employed a combination of Vector Autoregression (VAR) model and Graph Neural Networks (GNN) to systematically construct dynamic causal inference. Multiple classic classification algorithms were compared, including Random Forest, Logistic Regression, XGBoost, Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Gradient Boosting, and Multi Layer Perceptron (MLP). The SMOTE algorithm was used to undersample a small number of samples and employed Stratified K-fold Cross Validation. Results: This study included a total of 11,789 participants, including 6,334 females (53.73%) and 5,455 males (46.27%), with an average age of 65 years. Introduction of dynamic causal inference features has significantly improved the performance of almost all models. The area under the ROC curve of each model ranged from 0.78 to 0.83, indicating significant difference (P < 0.01). Among all the models, the Gradient Boosting model demonstrated the highest performance and stability. Model explanation and feature importance analysis generated model interpretation that illustrated significant contributors associated with risks of stroke. Conclusions and Relevance: This study proposes a stroke risk prediction method that combines dynamic causal inference with machine learning models, significantly improving prediction accuracy and revealing key health factors that affect stroke. The research results indicate that dynamic causal inference features have important value in predicting stroke risk, especially in capturing the impact of changes in health status over time on stroke risk. By further optimizing the model and introducing more variables, this study provides theoretical basis and practical guidance for future stroke prevention and intervention strategies.

While deep generative models have significantly advanced representation learning, they may inherit or amplify biases and fairness issues by encoding sensitive attributes alongside predictive features. Enforcing strict independence in disentanglement is often unrealistic when target and sensitive factors are naturally correlated. To address this challenge, we propose CAD-VAE (Correlation-Aware Disentangled VAE), which introduces a correlated latent code to capture the shared information between target and sensitive attributes. Given this correlated latent, our method effectively separates overlapping factors without extra domain knowledge by directly minimizing the conditional mutual information between target and sensitive codes. A relevance-driven optimization strategy refines the correlated code by efficiently capturing essential correlated features and eliminating redundancy. Extensive experiments on benchmark datasets demonstrate that CAD-VAE produces fairer representations, realistic counterfactuals, and improved fairness-aware image editing.

Interpretable machine learning models offer understandable reasoning behind their decision-making process, though they may not always match the performance of their black-box counterparts. This trade-off between interpretability and model performance has sparked discussions around the deployment of AI, particularly in critical applications where knowing the rationale of decision-making is essential for trust and accountability. In this study, we conduct a comparative analysis of several black-box and interpretable models, focusing on a specific NLP use case that has received limited attention: inferring ratings from reviews. Through this use case, we explore the intricate relationship between the performance and interpretability of different models. We introduce a quantitative score called Composite Interpretability (CI) to help visualize the trade-off between interpretability and performance, particularly in the case of composite models. Our results indicate that, in general, the learning performance improves as interpretability decreases, but this relationship is not strictly monotonic, and there are instances where interpretable models are more advantageous.

Despite their remarkable potential, Large Vision-Language Models (LVLMs) still face challenges with object hallucination, a problem where their generated outputs mistakenly incorporate objects that do not actually exist. Although most works focus on addressing this issue within the language-model backbone, our work shifts the focus to the image input source, investigating how specific image tokens contribute to hallucinations. Our analysis reveals a striking finding: a small subset of image tokens with high attention scores are the primary drivers of object hallucination. By removing these hallucinatory image tokens (only 1.5% of all image tokens), the issue can be effectively mitigated. This finding holds consistently across different models and datasets. Building on this insight, we introduce EAZY, a novel, training-free method that automatically identifies and Eliminates hAllucinations by Zeroing out hallucinatorY image tokens. We utilize EAZY for unsupervised object hallucination detection, achieving 15% improvement compared to previous methods. Additionally, EAZY demonstrates remarkable effectiveness in mitigating hallucinations while preserving model utility and seamlessly adapting to various LVLM architectures.

Recent inductive logic programming (ILP) approaches learn optimal hypotheses. An optimal hypothesis minimises a given cost function on the training data. There are many cost functions, such as minimising training error, textual complexity, or the description length of hypotheses. However, selecting an appropriate cost function remains a key question. To address this gap, we extend a constraint-based ILP system to learn optimal hypotheses for seven standard cost functions. We then empirically compare the generalisation error of optimal hypotheses induced under these standard cost functions. Our results on over 20 domains and 1000 tasks, including game playing, program synthesis, and image reasoning, show that, while no cost function consistently outperforms the others, minimising training error or description length has the best overall performance. Notably, our results indicate that minimising the size of hypotheses does not always reduce generalisation error.

In recent years, there has been increased interest in the design, training, and evaluation of end-to-end autonomous driving (AD) systems. One often overlooked aspect is the uncertainty of planned trajectories predicted by these systems, despite awareness of their own uncertainty being key to achieve safety and robustness. We propose to estimate this uncertainty by adapting loss prediction from the uncertainty quantification literature. To this end, we introduce a novel light-weight module, dubbed CATPlan, that is trained to decode motion and planning embeddings into estimates of the collision loss used to partially supervise end-to-end AD systems. During inference, these estimates are interpreted as collision risk. We evaluate CATPlan on the safety-critical, nerf-based, closed-loop benchmark NeuroNCAP and find that it manages to detect collisions with a $54.8\%$ relative improvement to average precision over a GMM-based baseline in which the predicted trajectory is compared to the forecasted trajectories of other road users. Our findings indicate that the addition of CATPlan can lead to safer end-to-end AD systems and hope that our work will spark increased interest in uncertainty quantification for such systems.

Object detection systems must reliably perceive objects of interest without being overly confident to ensure safe decision-making in dynamic environments. Filtering techniques based on out-of-distribution (OoD) detection are commonly added as an extra safeguard to filter hallucinations caused by overconfidence in novel objects. Nevertheless, evaluating YOLO-family detectors and their filters under existing OoD benchmarks often leads to unsatisfactory performance. This paper studies the underlying reasons for performance bottlenecks and proposes a methodology to improve performance fundamentally. Our first contribution is a calibration of all existing evaluation results: Although images in existing OoD benchmark datasets are claimed not to have objects within in-distribution (ID) classes (i.e., categories defined in the training dataset), around 13% of objects detected by the object detector are actually ID objects. Dually, the ID dataset containing OoD objects can also negatively impact the decision boundary of filters. These ultimately lead to a significantly imprecise performance estimation. Our second contribution is to consider the task of hallucination reduction as a joint pipeline of detectors and filters. By developing a methodology to carefully synthesize an OoD dataset that semantically resembles the objects to be detected, and using the crafted OoD dataset in the fine-tuning of YOLO detectors to suppress the objectness score, we achieve a 88% reduction in overall hallucination error with a combined fine-tuned detection and filtering system on the self-driving benchmark BDD-100K. Our code and dataset are available at: https://gricad-gitlab.univ-grenoble-alpes.fr/dnn-safety/m-hood.