If had debuted this year, William Shakespeare might have been accused of writing it with AI. A certain suspicious rhetorical device appears again and again in the play. It's in Act I, Scene ii: "The fault, dear Brutus, is not in our stars, but in ourselves." In Act III, Scene ii: "Not that I loved Caesar less, but that I loved Rome more." And later in that same scene: "I come to bury Caesar, not to praise him."

The latest interview in our series with the AAAI/SIGAI Doctoral Consortium participants features Thi Kieu Khanh Ho who is studying time-series anomaly detection. We found out more about her research, and what inspired her to study AI, and what she plans to work on next. Tell us a bit about your PhD -- where are you studying, and what is the topic of your research? I am doing my PhD at McGill University and Mila - Québec AI Institute, in the Department of Electrical and Computer Engineering, supervised by Professor Narges Armanfard. My research focuses on time-series anomaly detection, the problem of teaching AI systems to recognize when something unusual or abnormal is happening in complex, real-world data streams, without relying on large amounts of labeled examples.

We study how training data contributes to the emergence of toxic behaviors in large language models. Most prior work on reducing model toxicity adopts *reactive* approaches, such as fine-tuning pre-trained (and potentially toxic) models to align them with human values. In contrast, we propose a *proactive* approach--IF-Guide--that leverages influence functions to identify and suppress harmful tokens in the training data. To this end, we first show that standard influence functions are ineffective at discovering harmful training records. We then present a novel adaptation that measures token-level attributions from training data to model toxicity, along with techniques for selecting toxic training documents and a learning objective that can be integrated into both pre-training and fine-tuning. Moreover, IF-Guide does not rely on human-preference data, which is typically required by existing alignment methods. In our evaluation, we demonstrate that IF-Guide substantially reduces both explicit and implicit toxicity--by up to 10$\times$ compared to uncensored models, and up to 3$\times$ compared to baseline alignment methods such as DPO and RAD--across both pre-training and fine-tuning scenarios. IF-Guide is computationally efficient: a billion-parameter model is *not necessary* for computing influence scores; a million-parameter model--with 7.5$\times$ fewer parameters--can effectively serve as a proxy for identifying harmful data.

Mean Field Variational Inference (MFVI) is widely understood to underestimate posterior variance. By analysing conjugate Bayesian Linear Regression (BLR), we show that this characterization is incomplete: while MFVI underestimates the variance in parameter space, it can overestimate the predictive variance compared to the exact posterior. We show that if the MFVI posterior underestimates predictive variances in some directions, it necessarily overestimates them in others. Crucially, this overestimation occurs in directions where the training data concentrates. This leads to the surprising result that, for a test point drawn from the training distribution, MFVI's expected predictive variance exceeds that of the exact posterior. We demonstrate a pathological case of this effect, where the MFVI posterior fails to reduce predictive variance compared to the prior on in distribution data. We connect these results to the Cold Posterior Effect, arguing that varying the temperature can correct this overestimation, yielding predictions closer to those of the exact posterior. We validate our theory on synthetic and real-world regression tasks.

Large vision-language models (LVLMs) derive their capabilities from extensive training on vast corpora of visual and textual data.

Firms' algorithm development practices are often homogeneous. Whether firms train algorithms on similar data or rely on similar pre-trained models, the result is correlated predictions. In the context of personalized pricing, correlated algorithms can be viewed as a means to collude among competing firms, but whether or not this conduct is legal depends on the mechanisms of achieving collusion. We investigate the precise mechanisms through a formal game-theoretic model. Indeed, we find that (1) higher correlation diminishes consumer welfare and (2) as consumers become more price sensitive, firms are increasingly incentivized to compromise on the accuracy of their predictions in exchange for coordination. We demonstrate our theoretical results in a stylized empirical study where two firms compete using personalized pricing algorithms. Our results demonstrate a new mechanism for achieving collusion through correlation, which allows us to analyze its legal implications. Correlation through algorithms is a new frontier of anti-competitive behavior that is largely unconsidered by US antitrust law.

Foundation models trained on large-scale de-identified electronic health records (EHRs) hold promise for clinical applications. However, their capacity to memorize patient information raises important privacy concerns. In this work, we introduce a suite of black-box evaluation tests to assess privacy-related memorization risks in foundation models trained on structured EHR data. Our framework includes methods for probing memorization at both the embedding and generative levels, and aims to distinguish between model generalization and harmful memorization in clinically relevant settings. We contextualize memorization in terms of its potential to compromise patient privacy, particularly for vulnerable subgroups.

Out-of-distribution (OOD) 3D relighting requires novel view synthesis under unseen lighting conditions that differ significantly from the observed images. Existing relighting methods, which assume consistent light source distributions between training and testing, often degrade in OOD scenarios. We introduce MetaGS to tackle this challenge from two perspectives. First, we propose a meta-learning approach to train 3DGaussian splatting, which explicitly promotes learning generalizable Gaussian geometries and appearance attributes across diverse lighting conditions, even with biased training data. Second, we embed fundamental physical priors from the Blinn-Phong reflection model into Gaussian splatting, which enhances the decoupling of shading components and leads to more accurate 3D scene reconstruction. Results on both synthetic and real-world datasets demonstrate the effectiveness of MetaGS in challenging OOD relighting tasks, supporting efficient point-light relighting and generalizing well to unseen environment lighting maps.

Safe deployment of machine learning (ML) models in safety-critical domains such as medical imaging requires detecting inputs with characteristics not seen during training, known as out-of-distribution (OOD) detection, to prevent unreliable predictions. Effective OOD detection after deployment could benefit from access to the training data, enabling direct comparison between test samples and the training data distribution to identify differences. State-of-the-art OOD detection methods, however, either discard the training data after deployment or assume that test samples and training data are centrally stored together, an assumption that rarely holds in real-world settings. This is because shipping the training data with the deployed model is usually impossible due to the size of training databases, as well as proprietary or privacy constraints. We introduce the Isolation Network, an OOD detection framework that quantifies the difficulty of separating a target test sample from the training data by solving a binary classification task.

We thereby identified the offset effect, a phenomenon characterized by two key findings: (1) verbatim memorization is most strongly triggered by short prefixes drawn from the beginning of the context window, with memorization decreasing counterintuitively as prefix length increases; and (2) a sharp decline in verbatim recall when prefix begins offset from the initial tokens of the context window. We attribute this to positional fragility: models rely disproportionately on the earliest tokens in their context window as retrieval anchors, making them sensitive to even slight shifts. We further observe that when the model fails to retrieve memorized content, it often produces degenerated text. Leveraging these findings, we show that shifting sensitive data deeper into the context window suppresses both extractable memorization and degeneration. Our results suggest that positional offset is a critical and previously overlooked axis for evaluating memorization risks, since prior work implicitly assumed uniformity by probing only from the beginning of documents or training sequences.