Acoustic telemetry data plays a vital role in understanding the behaviour and movement of aquatic animals. However, these datasets, which often consist of millions of individual data points, frequently contain anomalous movements that pose significant challenges. Traditionally, anomalous movements are identified either manually or through basic statistical methods, approaches that are time-consuming and prone to high rates of unidentified anomalies in large datasets. This study focuses on the development of automated classifiers for a large telemetry dataset comprising detections from fifty acoustically tagged dusky kob monitored in the Breede Estuary, South Africa. Using an array of 16 acoustic receivers deployed throughout the estuary between 2016 and 2021, we collected over three million individual data points. We present detailed guidelines for data pre-processing, resampling strategies, labelling process, feature engineering, data splitting methodologies, and the selection and interpretation of machine learning and deep learning models for anomaly detection. Among the evaluated models, neural networks autoencoder (NN-AE) demonstrated superior performance, aided by our proposed threshold-finding algorithm. NN-AE achieved a high recall with no false normal (i.e., no misclassifications of anomalous movements as normal patterns), a critical factor in ensuring that no true anomalies are overlooked. In contrast, other models exhibited false normal fractions exceeding 0.9, indicating they failed to detect the majority of true anomalies; a significant limitation for telemetry studies where undetected anomalies can distort interpretations of movement patterns. While the NN-AE's performance highlights its reliability and robustness in detecting anomalies, it faced challenges in accurately learning normal movement patterns when these patterns gradually deviated from anomalous ones.

While the community has designed various defenses to counter the threat of poisoning attacks in Federated Learning (FL), there are no guidelines for evaluating these defenses. These defenses are prone to subtle pitfalls in their experimental setups that lead to a false sense of security, rendering them unsuitable for practical deployment. In this paper, we systematically understand, identify, and provide a better approach to address these challenges. First, we design a comprehensive systemization of FL defenses along three dimensions: i) how client updates are processed, ii) what the server knows, and iii) at what stage the defense is applied. Next, we thoroughly survey 50 top-tier defense papers and identify the commonly used components in their evaluation setups. Based on this survey, we uncover six distinct pitfalls and study their prevalence. For example, we discover that around 30% of these works solely use the intrinsically robust MNIST dataset, and 40% employ simplistic attacks, which may inadvertently portray their defense as robust. Using three representative defenses as case studies, we perform a critical reevaluation to study the impact of the identified pitfalls and show how they lead to incorrect conclusions about robustness. We provide actionable recommendations to help researchers overcome each pitfall.

Detailed image captioning is essential for tasks like data generation and aiding visually impaired individuals. High-quality captions require a balance between precision and recall, which remains challenging for current multimodal large language models (MLLMs). In this work, we hypothesize that this limitation stems from weakening and increasingly noisy visual attention as responses lengthen. To address this issue, we propose SPARC (Selective Progressive Attention ReCalibration), a training-free method that enhances the contribution of visual tokens during decoding. SPARC is founded on three key observations: (1) increasing the influence of all visual tokens reduces recall; thus, SPARC selectively amplifies visual tokens; (2) as captions lengthen, visual attention becomes noisier, so SPARC identifies critical visual tokens by leveraging attention differences across time steps; (3) as visual attention gradually weakens, SPARC reinforces it to preserve its influence. Our experiments, incorporating both automated and human evaluations, demonstrate that existing methods improve the precision of MLLMs at the cost of recall. In contrast, our proposed method enhances both precision and recall with minimal computational overhead.

The goal of inductive logic programming (ILP) is to find a set of logical rules that generalises training examples and background knowledge. We introduce an ILP approach that identifies pointless rules. A rule is pointless if it contains a redundant literal or cannot discriminate against negative examples. We show that ignoring pointless rules allows an ILP system to soundly prune the hypothesis space. Our experiments on multiple domains, including visual reasoning and game playing, show that our approach can reduce learning times by 99% whilst maintaining predictive accuracies.

Generating semantically coherent text requires a robust internal representation of linguistic structures, which traditional embedding techniques often fail to capture adequately. A novel approach, Latent Lexical Projection (LLP), is introduced to refine lexical representations through a structured transformation into a latent space, thereby enhancing the alignment between input embeddings and their contextual meanings. The method integrates an optimized projection mechanism within an existing language model architecture, enabling more accurate token selection while maintaining syntactic integrity. Evaluations across multiple benchmarks indicate a reduction in perplexity and an increase in BLEU scores, suggesting improvements in predictive accuracy and fluency. The analysis of lexical diversity reveals a more varied vocabulary in generated text, addressing common issues of redundancy and repetitive phrase structures. Further assessments of entropy distributions demonstrate a decline in uncertainty during decoding, reflecting enhanced confidence in word selection. Additionally, long-range dependency retention exhibits measurable gains, with increased classification accuracy at extended token distances. Computational efficiency remains within manageable constraints, despite the added projection mechanism, highlighting the practicality of LLP for integration into existing architectures.

Despite the growing interest in human-AI decision making, experimental studies with domain experts remain rare, largely due to the complexity of working with domain experts and the challenges in setting up realistic experiments. In this work, we conduct an in-depth collaboration with radiologists in prostate cancer diagnosis based on MRI images. Building on existing tools for teaching prostate cancer diagnosis, we develop an interface and conduct two experiments to study how AI assistance and performance feedback shape the decision making of domain experts. In Study 1, clinicians were asked to provide an initial diagnosis (human), then view the AI's prediction, and subsequently finalize their decision (human-AI team). In Study 2 (after a memory wash-out period), the same participants first received aggregated performance statistics from Study 1, specifically their own performance, the AI's performance, and their human-AI team performance, and then directly viewed the AI's prediction before making their diagnosis (i.e., no independent initial diagnosis). These two workflows represent realistic ways that clinical AI tools might be used in practice, where the second study simulates a scenario where doctors can adjust their reliance and trust on AI based on prior performance feedback. Our findings show that, while human-AI teams consistently outperform humans alone, they still underperform the AI due to under-reliance, similar to prior studies with crowdworkers. Providing clinicians with performance feedback did not significantly improve the performance of human-AI teams, although showing AI decisions in advance nudges people to follow AI more. Meanwhile, we observe that the ensemble of human-AI teams can outperform AI alone, suggesting promising directions for human-AI collaboration.

Unsupervised anomaly detection is a challenging task. Autoencoders (AEs) or generative models are often employed to model the data distribution of normal inputs and subsequently identify anomalous, out-of-distribution inputs by high reconstruction error or low likelihood, respectively. However, AEs may generalize and achieve small reconstruction errors on abnormal inputs. We propose a decoupled training approach for anomaly detection that both an AE and a likelihood model trained with noise contrastive estimation (NCE). After training the AE, NCE estimates a probability density function, to serve as the anomaly score, on the joint space of the AE's latent representation combined with features of the reconstruction quality. To further reduce the false negative rate in NCE we systematically varying the reconstruction features to augment the training and optimize the contrastive Gaussian noise distribution. Experimental assessments on multiple benchmark datasets demonstrate that the proposed approach matches the performance of prevalent state-of-the-art anomaly detection algorithms.

The number of applications in Google Play has increased dramatically in recent years. On Google Play, users can write detailed reviews and rate apps, with these ratings significantly influencing app success and download numbers. Reviews often include notable information like feature requests, which are valuable for software maintenance. Users can update their reviews and ratings anytime. Studies indicate that apps with ratings below three stars are typically avoided by potential users. Since 2013, Google Play has allowed developers to respond to user reviews, helping resolve issues and potentially boosting overall ratings and download rates. However, responding to reviews is time-consuming, and only 13% to 18% of developers engage in this practice. To address this challenge, we propose a method to prioritize reviews based on response priority. We collected and preprocessed review data, extracted both textual and semantic features, and assessed their impact on the importance of responses. We labelled reviews as requiring a response or not and trained four different machine learning models to prioritize them. We evaluated the models performance using metrics such as F1-Score, Accuracy, Precision, and Recall. Our findings indicate that the XGBoost model is the most effective for prioritizing reviews needing a response.

Sampling-based search, a simple paradigm for utilizing test-time compute, involves generating multiple candidate responses and selecting the best one--typically by having models self-verify each response for correctness. In this paper, we study the scaling trends governing sampling-based search. Among our findings is that simply scaling up a minimalist implementation of sampling-based search, using only random sampling and direct self-verification, provides a practical inference method that, for example, elevates the reasoning capabilities of Gemini v1.5 Pro above that of o1-Preview on popular benchmarks. We partially attribute the scalability of sampling-based search to a phenomenon of implicit scaling, where sampling a larger pool of responses in turn improves self-verification accuracy. We further identify two useful principles for improving self-verification capabilities with test-time compute: (1) comparing across responses provides helpful signals about the locations of errors and hallucinations, and (2) different model output styles are useful for different contexts--chains of thought are useful for reasoning but harder to verify. We also find that, though accurate verification can be elicited, frontier models demonstrate remarkably weak out-of-box verification capabilities and introduce a benchmark to measure progress on these deficiencies.

Tool-augmented large language models (LLMs) are often trained on datasets of query-response pairs, which embed the ability to use tools or APIs directly into the parametric knowledge of LLMs. Tool-augmented LLMs need the ability to forget learned tools due to security vulnerabilities, privacy regulations, or tool deprecations. However, ``tool unlearning'' has not been investigated in unlearning literature. We introduce this novel task, which requires addressing distinct challenges compared to traditional unlearning: knowledge removal rather than forgetting individual samples, the high cost of optimizing LLMs, and the need for principled evaluation metrics. To bridge these gaps, we propose ToolDelete, the first approach for unlearning tools from tool-augmented LLMs. It implements three key properties to address the above challenges for effective tool unlearning and introduces a new membership inference attack (MIA) model for effective evaluation. Extensive experiments on multiple tool learning datasets and tool-augmented LLMs show that ToolDelete effectively unlearns randomly selected tools, while preserving the LLM's knowledge on non-deleted tools and maintaining performance on general tasks.