Since the advent of machine learning, interpretability has remained a persistent challenge, becoming increasingly urgent as generative models support high-stakes applications in drug and material discovery. Recent advances in large language model (LLM) architectures have yielded chemistry language models (CLMs) with impressive capabilities in molecular property prediction and molecular generation. However, how these models internally represent chemical knowledge remains poorly understood. In this work, we extend sparse autoencoder techniques to uncover and examine interpretable features within CLMs. Applying our methodology to the Foundation Models for Materials (FM4M) SMI-TED chemistry foundation model, we extract semantically meaningful latent features and analyse their activation patterns across diverse molecular datasets. Our findings reveal that these models encode a rich landscape of chemical concepts. We identify correlations between specific latent features and distinct domains of chemical knowledge, including structural motifs, physicochemical properties, and pharmacological drug classes. Our approach provides a generalisable framework for uncovering latent knowledge in chemistry-focused AI systems. This work has implications for both foundational understanding and practical deployment; with the potential to accelerate computational chemistry research.

We propose an interpretable deep competing risks model called the Deep Kernel Aalen-Johansen (DKAJ) estimator, which generalizes the classical Aalen-Johansen nonparametric estimate of cumulative incidence functions (CIFs). Each data point (e.g., patient) is represented as a weighted combination of clusters. If a data point has nonzero weight only for one cluster, then its predicted CIFs correspond to those of the classical Aalen-Johansen estimator restricted to data points from that cluster. These weights come from an automatically learned kernel function that measures how similar any two data points are. On four standard competing risks datasets, we show that DKAJ is competitive with state-of-the-art baselines while being able to provide visualizations to assist model interpretation.

Time series forecasting has applications across domains and industries, especially in healthcare, but the technical expertise required to analyze data, build models, and interpret results can be a barrier to using these techniques. This article presents a web platform that makes the process of analyzing and plotting data, training forecasting models, and interpreting and viewing results accessible to researchers and clinicians. Users can upload data and generate plots to showcase their variables and the relationships between them. The platform supports multiple forecasting models and training techniques which are highly customizable according to the user's needs. Additionally, recommendations and explanations can be generated from a large language model that can help the user choose appropriate parameters for their data and understand the results for each model. The goal is to integrate this platform into learning health systems for continuous data collection and inference from clinical pipelines.

Continual learning constrains models to learn new tasks over time without forgetting what they have already learned. A key challenge in this setting is catastrophic forgetting, where learning new information causes the model to lose its performance on previous tasks. Recently, explainable AI has been proposed as a promising way to better understand and reduce forgetting. In particular, self-explainable models are useful because they generate explanations during prediction, which can help preserve knowledge. However, most existing explainable approaches use post-hoc explanations or require additional memory for each new task, resulting in limited scalability. In this work, we introduce CIP-Net, an exemplar-free self-explainable prototype-based model designed for continual learning. CIP-Net avoids storing past examples and maintains a simple architecture, while still providing useful explanations and strong performance. We demonstrate that CIP-Net achieves state-of-the-art performances compared to previous exemplar-free and self-explainable methods in both task-and class-incremental settings, while bearing significantly lower memory-related overhead. This makes it a practical and interpretable solution for continual learning.

Medical decision-making makes frequent use of algorithms that combine risk equations with rules, providing clear and standardized treatment pathways. Symbolic regression (SR) traditionally limits its search space to continuous function forms and their parameters, making it difficult to model this decision-making. However, due to its ability to derive data-driven, interpretable models, SR holds promise for developing data-driven clinical risk scores. To that end we introduce Brush, an SR algorithm that combines decision-tree-like splitting algorithms with non-linear constant optimization, allowing for seamless integration of rule-based logic into symbolic regression and classification models. Brush achieves Pareto-optimal performance on SRBench, and was applied to recapitulate two widely used clinical scoring systems, achieving high accuracy and interpretable models. Compared to decision trees, random forests, and other SR methods, Brush achieves comparable or superior predictive performance while producing simpler models.

Contrastive learning has emerged as a powerful method in deep learning, excelling at learning effective representations through contrasting samples from different distributions. However, dimensional collapse, where embeddings converge into a lower-dimensional space, poses a significant challenge, especially in semi-supervised and self-supervised setups. In this paper, we first identify a critical learning-rate threshold, beyond which standard contrastive losses converge to collapsed solutions. Building on these insights, we propose CLOP, a novel semi-supervised loss function designed to prevent dimensional collapse by promoting the formation of orthogonal linear subspaces among class embeddings. Through extensive experiments on real and synthetic datasets, we demonstrate that CLOP improves performance in image classification and object detection tasks while also exhibiting greater stability across different learning rates and batch sizes.

Nonnegative matrix factorization (NMF) is a widely used tool for learning parts-based, low-dimensional representations of nonnegative data, with applications in vision, text, and bioinformatics. In clustering applications, orthogonal NMF (ONMF) variants further impose (approximate) orthogonality on the representation matrix so that its rows behave like soft cluster indicators. Existing algorithms, however, are typically derived from optimization viewpoints and do not explicitly exploit the conic geometry induced by NMF: data points lie in a convex cone whose extreme rays encode fundamental directions or "topics". In this work we revisit NMF from this geometric perspective and propose Cone Collapse, an algorithm that starts from the full nonnegative orthant and iteratively shrinks it toward the minimal cone generated by the data. We prove that, under mild assumptions on the data, Cone Collapse terminates in finitely many steps and recovers the minimal generating cone of $\mathbf{X}^\top$ . Building on this basis, we then derive a cone-aware orthogonal NMF model (CC-NMF) by applying uni-orthogonal NMF to the recovered extreme rays. Across 16 benchmark gene-expression, text, and image datasets, CC-NMF consistently matches or outperforms strong NMF baselines-including multiplicative updates, ANLS, projective NMF, ONMF, and sparse NMF-in terms of clustering purity. These results demonstrate that explicitly recovering the data cone can yield both theoretically grounded and empirically strong NMF-based clustering methods.

Foundation models have shown remarkable success in fitting biological visual systems; however, their black-box nature inherently limits their utility for understanding brain function. Here, we peek inside a SOTA foundation model of neural activity (Wang et al., 2025) as a physiologist might, characterizing each 'neuron' based on its temporal response properties to parametric stimuli. We analyze how different stimuli are represented in neural activity space by building decoding manifolds, and we analyze how different neurons are represented in stimulus-response space by building neural encoding manifolds. We find that the different processing stages of the model (i.e., the feedforward encoder, recurrent, and readout modules) each exhibit qualitatively different representational structures in these manifolds. The recurrent module shows a jump in capabilities over the encoder module by 'pushing apart' the representations of different temporal stimulus patterns; while the readout module achieves biological fidelity by using numerous specialized feature maps rather than biologically plausible mechanisms. Overall, we present this work as a study of the inner workings of a prominent neural foundation model, gaining insights into the biological relevance of its internals through the novel analysis of its neurons' joint temporal response patterns.

Traffic forecasting task is significant to modern urban management. Recently, there is growing attention on large-scale forecasting, as it better reflects the complexity of real-world traffic networks. However, existing models often exhibit quadratic computational complexity, making them impractical for large-scale real-world scenarios. In this paper, we propose a novel framework, Hierarchical Spatio-Temporal Mixer (HSTMixer), which leverages an all-MLP architecture for efficient and effective large-scale traffic forecasting. HSTMixer employs a hierarchical spatiotemporal mixing block to extract multi-resolution features through bottom-up aggregation and top-down propagation. Furthermore, an adaptive region mixer generates transformation matrices based on regional semantics, enabling our model to dynamically capture evolving spatiotemporal patterns for different regions. Extensive experiments conducted on four large-scale real-world datasets demonstrate that the proposed method not only achieves state-of-the-art performance but also exhibits competitive computational efficiency.

Despite rapid progress in LLM agents, performance on long-horizon, tool-using tasks remains fragile. To better understand this fragility, we ask a simple question: \emph{do all actions contribute equally to failure?} Analyzing execution traces on $τ$-Bench (Airline/Retail) and SWE-Bench Verified, we decompose trajectories into \emph{mutating} (environment-changing) vs.\ non-mutating steps and formalize \emph{decisive deviations}, earliest action, level divergences that flip success to failure. A logistic regression reveals that each additional deviation in a mutating action reduces the odds of success by upto $92\%$ on Airline and upto $96\%$ on Retail for SoTA models. In contrast, deviations in non-mutating actions have little to no effect. Errors also grow with context length as agents drift from role and act on stale constraints. Motivated by these observations, we introduce \cm{}, a model-agnostic, gradient-free, test-time safeguard that (i) adds mutation-gated verification, (ii) injects \emph{Targeted Reflection} before mutating steps, and (iii) performs block-based context cleaning. \cm{} delivers consistent gains, e.g., Qwen3-Thinking: +28\% \emph{relative} on Airline, +11\% on Retail, and +7\% on SWE-Bench Verified; Claude: +9\%/+7\%. We further identify ceiling effects in $τ$-Bench, where annotation errors and underspecified tasks artificially cap model performance. To address this, we release $τ$-Bench Verified, which restores benchmark headroom through targeted revisions. Our results argue for action-level analysis, targeted safeguards, and reliable evaluations as prerequisites for robust multi-turn agents.